Search results for: image processing

Authors: Yangzhi Li

Abstract:

Network transfer of information and performance customization is now a viable method of digital industrial production in the era of Industry 4.0. Robot platforms and network platforms have grown more important in digital design and construction. The pressing need for novel building techniques is driven by the growing labor scarcity problem and increased awareness of construction safety. Robotic approaches in construction research are regarded as an extension of operational and production tools. Several technological theories related to robot autonomous recognition, which include high-performance computing, physical system modeling, extensive sensor coordination, and dataset deep learning, have not been explored using intelligent construction. Relevant transdisciplinary theory and practice research still has specific gaps. Optimizing high-performance computing and autonomous recognition visual guidance technologies improves the robot's grasp of the scene and capacity for autonomous operation. Intelligent vision guidance technology for industrial robots has a serious issue with camera calibration, and the use of intelligent visual guiding and identification technologies for industrial robots in industrial production has strict accuracy requirements. It can be considered that visual recognition systems have challenges with precision issues. In such a situation, it will directly impact the effectiveness and standard of industrial production, necessitating a strengthening of the visual guiding study on positioning precision in recognition technology. To best facilitate the handling of complicated components, an approach for the visual recognition of parts utilizing machine learning algorithms is proposed. This study will identify the position of target components by detecting the information at the boundary and corner of a dense point cloud and determining the aspect ratio in accordance with the guidelines for the modularization of building components. To collect and use components, operational processing systems assign them to the same coordinate system based on their locations and postures. The RGB image's inclination detection and the depth image's verification will be used to determine the component's present posture. Finally, a virtual environment model for the robot's obstacle-avoidance route will be constructed using the point cloud information.

Keywords: robotic construction, robotic assembly, visual guidance, machine learning

Procedia PDF Downloads 86

Authors: V. E. Messerle, A. L. Mosse, O. A. Lavrichshev, A. N. Nikonchuk, A. B. Ustimenko

Abstract:

One of the most serious environmental problems today is pollution by biomedical waste (BMW), which in most cases has undesirable properties such as toxicity, carcinogenicity, mutagenicity, fire. Sanitary and hygienic survey of typical solid BMW, made in Belarus, Kazakhstan, Russia and other countries shows that their risk to the environment is significantly higher than that of most chemical wastes. Utilization of toxic BMW requires use of the most universal methods to ensure disinfection and disposal of any of their components. Such technology is a plasma technology of BMW processing. To implement this technology a thermodynamic analysis of the plasma processing of BMW was fulfilled and plasma-box furnace was developed. The studies have been conducted on the example of the processing of bone. To perform thermodynamic calculations software package Terra was used. Calculations were carried out in the temperature range 300 - 3000 K and a pressure of 0.1 MPa. It is shown that the final products do not contain toxic substances. From the organic mass of BMW synthesis gas containing combustible components 77.4-84.6% was basically produced, and mineral part consists mainly of calcium oxide and contains no carbon. Degree of gasification of carbon reaches 100% by the temperature 1250 K. Specific power consumption for BMW processing increases with the temperature throughout its range and reaches 1 kWh/kg. To realize plasma processing of BMW experimental installation with DC plasma torch of 30 kW power was developed. The experiments allowed verifying the thermodynamic calculations. Wastes are packed in boxes weighing 5-7 kg. They are placed in the box furnace. Under the influence of air plasma flame average temperature in the box reaches 1800 OC, the organic part of the waste is gasified and inorganic part of the waste is melted. The resulting synthesis gas is continuously withdrawn from the unit through the cooling and cleaning system. Molten mineral part of the waste is removed from the furnace after it has been stopped. Experimental studies allowed determining operating modes of the plasma box furnace, the exhaust gases was analyzed, samples of condensed products were assembled and their chemical composition was determined. Gas at the outlet of the plasma box furnace has the following composition (vol.%): CO - 63.4, H2 - 6.2, N2 - 29.6, S - 0.8. The total concentration of synthesis gas (CO + H2) is 69.6%, which agrees well with the thermodynamic calculation. Experiments confirmed absence of the toxic substances in the final products.

Keywords: biomedical waste, box furnace, plasma torch, processing, synthesis gas

Procedia PDF Downloads 525

Authors: Rios A. S., Hild F., Deus E. P., Aimedieu P., Benallal A.

Abstract:

The damage micromechanisms of chopped strand mats manufactured by compression of Brazilian coconut fiber and coconut fibers in different external conditions (chemical treatment) were used in this study. Mechanical analysis testing uniaxial traction were used with Digital Image Correlation (DIC). The images captured during the tensile test in the coconut fibers and coconut fiber mats showed an uncertainty of measurement in order centipixels. The initial modulus (modulus of elasticity) and tensile strength decreased with increasing diameter for the four conditions of coconut fibers. The DIC showed heterogeneous deformation fields for coconut fibers and mats and the displacement fields showed the rupture process of coconut fiber. The determination of poisson’s ratio of the mat was performed through of transverse and longitudinal deformations found in the elastic region.

Keywords: coconut fiber, mechanical behavior, digital image correlation, micromechanism

Procedia PDF Downloads 459

Authors: Jalal Karam

Abstract:

The methods of Fourier, Laplace, and Wavelet Transforms provide transfer functions and relationships between the input and the output signals in linear time invariant systems. This paper shows the equivalence among these three methods and in each case presenting an application of the appropriate (Fourier, Laplace or Wavelet) to the convolution theorem. In addition, it is shown that the same holds for a direct integration method. The Biorthogonal wavelets Bior3.5 and Bior3.9 are examined and the zeros distribution of their polynomials associated filters are located. This paper also presents the significance of utilizing wavelets as effective tools in processing speech signals for common multimedia applications in general, and for recognition and compression in particular. Theoretically and practically, wavelets have proved to be effective and competitive. The practical use of the Continuous Wavelet Transform (CWT) in processing and analysis of speech is then presented along with explanations of how the human ear can be thought of as a natural wavelet transformer of speech. This generates a variety of approaches for applying the (CWT) to many paradigms analysing speech, sound and music. For perception, the flexibility of implementation of this transform allows the construction of numerous scales and we include two of them. Results for speech recognition and speech compression are then included.

Keywords: continuous wavelet transform, biorthogonal wavelets, speech perception, recognition and compression

Procedia PDF Downloads 416

Authors: Sajeeha Ansar, Asad Ali Safi, Sheikh Ziauddin, Ahmad R. Shahid, Faraz Ahsan

Abstract:

Gliomas are most challenging and aggressive type of tumors which appear in different sizes, locations, and scattered boundaries. CNN is most efficient deep learning approach with outstanding capability of solving image analysis problems. A fully automatic deep learning based 2D-CNN model for brain tumor segmentation is presented in this paper. We used small convolution filters (3 x 3) to make architecture deeper. We increased convolutional layers for efficient learning of complex features from large dataset. We achieved better results by pushing convolutional layers up to 16 layers for HGG model. We achieved reliable and accurate results through fine-tuning among dataset and hyper-parameters. Pre-processing of this model includes generation of brain pipeline, intensity normalization, bias correction and data augmentation. We used the BRATS-2015, and Dice Similarity Coefficient (DSC) is used as performance measure for the evaluation of the proposed method. Our method achieved DSC score of 0.81 for complete, 0.79 for core, 0.80 for enhanced tumor regions. However, these results are comparable with methods already implemented 2D CNN architecture.

Keywords: brain tumor segmentation, convolutional neural networks, deep learning, HGG

Procedia PDF Downloads 256

Authors: Syed Basit Ali Shah Bukhari, Yan Qiang, Saad Abdul Rauf, Syed Saqlaina Bukhari

Abstract:

Text Classification is the methodology to classify any given text into the respective category from a given set of categories. It is highly important and vital to use proper set of pre-processing , feature selection and classification techniques to achieve this purpose. In this paper we have used different ensemble techniques along with variance in feature selection parameters to see the change in overall accuracy of the result and also on some other individual class based features which include precision value of each individual category of the text. After subjecting our data through pre-processing and feature selection techniques , different individual classifiers were tested first and after that classifiers were combined to form ensembles to increase their accuracy. Later we also studied the impact of decreasing the classification categories on over all accuracy of data. Text classification is highly used in sentiment analysis on social media sites such as twitter for realizing people’s opinions about any cause or it is also used to analyze customer’s reviews about certain products or services. Opinion mining is a vital task in data mining and text categorization is a back-bone to opinion mining.

Keywords: Natural Language Processing, Ensemble Classifier, Bagging Classifier, AdaBoost

Procedia PDF Downloads 232

Authors: Mohammad Ghahramani, Fahimeh Saei Manesh

Abstract:

Winning a soccer game is based on thorough and deep analysis of the ongoing match. On the other hand, giant gambling companies are in vital need of such analysis to reduce their loss against their customers. In this research work, we perform deep, real-time analysis on every soccer match around the world that distinguishes our work from others by focusing on particular seasons, teams and partial analytics. Our contributions are presented in the platform called “Analyst Masters.” First, we introduce various sources of information available for soccer analysis for teams around the world that helped us record live statistical data and information from more than 50,000 soccer matches a year. Our second and main contribution is to introduce our proposed in-play performance evaluation. The third contribution is developing new features from stable soccer matches. The statistics of soccer matches and their odds before and in-play are considered in the image format versus time including the halftime. Local Binary patterns, (LBP) is then employed to extract features from the image. Our analyses reveal incredibly interesting features and rules if a soccer match has reached enough stability. For example, our “8-minute rule” implies if 'Team A' scores a goal and can maintain the result for at least 8 minutes then the match would end in their favor in a stable match. We could also make accurate predictions before the match of scoring less/more than 2.5 goals. We benefit from the Gradient Boosting Trees, GBT, to extract highly related features. Once the features are selected from this pool of data, the Decision trees decide if the match is stable. A stable match is then passed to a post-processing stage to check its properties such as betters’ and punters’ behavior and its statistical data to issue the prediction. The proposed method was trained using 140,000 soccer matches and tested on more than 100,000 samples achieving 98% accuracy to select stable matches. Our database from 240,000 matches shows that one can get over 20% betting profit per month using Analyst Masters. Such consistent profit outperforms human experts and shows the inefficiency of the betting market. Top soccer tipsters achieve 50% accuracy and 8% monthly profit in average only on regional matches. Both our collected database of more than 240,000 soccer matches from 2012 and our algorithm would greatly benefit coaches and punters to get accurate analysis.

Keywords: soccer, analytics, machine learning, database

Procedia PDF Downloads 238

Authors: Nougrara Zineb

Abstract:

The unique structure of the blood vessels in the retina has been used for biometric identification. The retina blood vessel pattern is a unique pattern in each individual and it is almost impossible to forge that pattern in a false individual. The retina biometrics’ advantages include high distinctiveness, universality, and stability overtime of the blood vessel pattern. Once the creases have been extracted from the images, a registration stage is necessary, since the position of the retinal vessel structure could change between acquisitions due to the movements of the eye. Image registration consists of following steps: Feature detection, feature matching, transform model estimation and image resembling and transformation. In this paper, we present an algorithm of registration; it is based on the characterization of retinal feature points. For experiments, retinal images from the DRIVE database have been tested. The proposed methodology achieves good results for registration in general.

Keywords: fovea, optic disc, registration, retinal images

Procedia PDF Downloads 266

Authors: Chirawat Wattanapanich, Hong Wei

Abstract:

This study presents new gait representations for improving gait recognition accuracy on cross gait appearances, such as normal walking, wearing a coat and carrying a bag. Based on the Gait Energy Image (GEI), two ideas are implemented to generate new gait representations. One is to append lower knee regions to the original GEI, and the other is to apply convolutional operations to the GEI and its variants. A set of new gait representations are created and used for training multi-class Support Vector Machines (SVMs). Tests are conducted on the CASIA dataset B. Various combinations of the gait representations with different convolutional kernel size and different numbers of kernels used in the convolutional processes are examined. Both the entire images as features and reduced dimensional features by Principal Component Analysis (PCA) are tested in gait recognition. Interestingly, both new techniques, appending the lower knee regions to the original GEI and convolutional GEI, can significantly contribute to the performance improvement in the gait recognition. The experimental results have shown that the average recognition rate can be improved from 75.65% to 87.50%.

Keywords: convolutional image, lower knee, gait

Procedia PDF Downloads 202

Authors: Cherry Yieng Siang Ling, Joong Hee Lee, Myung Hwan Yun

Abstract:

The quality of a product to be usable has become the basic requirement in consumer’s perspective while failing the requirement ends up the customer from not using the product. Identifying usability issues from analyzing quantitative and qualitative data collected from usability testing and evaluation activities aids in the process of product design, yet the lack of studies and researches regarding analysis methodologies in qualitative text data of usability field inhibits the potential of these data for more useful applications. While the possibility of analyzing qualitative text data found with the rapid development of data analysis studies such as natural language processing field in understanding human language in computer, and machine learning field in providing predictive model and clustering tool. Therefore, this research aims to study the application capability of text processing algorithm in analysis of qualitative text data collected from usability activities. This research utilized datasets collected from LG neckband headset usability experiment in which the datasets consist of headset survey text data, subject’s data and product physical data. In the analysis procedure, which integrated with the text-processing algorithm, the process includes training of comments onto vector space, labeling them with the subject and product physical feature data, and clustering to validate the result of comment vector clustering. The result shows 'volume and music control button' as the usability feature that matches best with the cluster of comment vectors where centroid comments of a cluster emphasized more on button positions, while centroid comments of the other cluster emphasized more on button interface issues. When volume and music control buttons are designed separately, the participant experienced less confusion, and thus, the comments mentioned only about the buttons' positions. While in the situation where the volume and music control buttons are designed as a single button, the participants experienced interface issues regarding the buttons such as operating methods of functions and confusion of functions' buttons. The relevance of the cluster centroid comments with the extracted feature explained the capability of text processing algorithms in analyzing qualitative text data from usability testing and evaluations.

Keywords: usability, qualitative data, text-processing algorithm, natural language processing

Procedia PDF Downloads 285

Authors: Edwin Hidayat, Redi Yulianto, Disi Hanafiah

Abstract:

On the one hand, parameters which are used for determining the number of lane on the new road construction are average annual average daily traffic (AADT) and peak hour factor (K-factor). On the other hand, the value of K-factor listed in the guidelines and manual for road planning in Indonesia is a value of adoption or adaptation from foreign guidelines or manuals. Thus, the value is less suitable for Indonesian condition due to differences in road conditions, vehicle type, and driving behavior. The purpose of this study is to provide an example on how to determine k-factor values at a road segment with particular conditions in north coast road, West Java. The methodology is started with collecting traffic volume data for 24 hours over 365 days using PLATO (Automated Traffic Counter) with the approach of video image processing. Then, the traffic volume data is divided into per hour and analyzed by comparing the peak traffic volume in the 30th hour (or other) with the AADT in the same year. The analysis has resulted that for the 30th peak hour the K-factor is 0.97. This value can be used for planning road geometry or evaluating the road capacity performance for the 4/2D interurban road.

Keywords: road geometry, K-factor, annual average daily traffic, north coast road

Procedia PDF Downloads 161

Authors: Duckyong Kim, Jong Kang Park, Jong Tae Kim

Abstract:

DOA (Direction of Arrival) estimation is an important method in array signal processing and has a wide range of applications such as direction finding, beam forming, and so on. In this paper, we briefly introduce the MUSIC (Multiple Signal Classification) Algorithm, one of DOA estimation methods for analyzing several targets. Then we apply the MUSIC algorithm to the two-dimensional antenna array to analyze DOA estimation in 3D space through MATLAB simulation. We also analyze the design factors that can affect the accuracy of DOA estimation through simulation, and proceed with further consideration on how to apply the system.

Keywords: DOA estimation, MUSIC algorithm, spatial spectrum, array signal processing

Procedia PDF Downloads 379

Authors: Khanh D.Do, Loi V.Nguyen, Thanh N.Nguyen, Thang M.Nguyen, Vu T.Tran

Abstract:

Along with the development of Secondary surveillance radar (SSR) in air traffic surveillance systems, the Multipath phenomena has always been a noticeable problem. This following article discusses the geometrical aspect and power aspect of the Multipath interference caused by reflection in SSR and proposes a method to deal with these unwanted multipath targets (ghosts) by false-target position predicting and adaptive target suppressing. A field-experiment example is mentioned at the end of the article to demonstrate the efficiency of this measure.

Keywords: multipath, secondary surveillance radar, digital signal processing, reflection

Procedia PDF Downloads 162

Authors: Su Yin Chiam, Qing Xin Zhang, Jaehoon Chung

Abstract:

Complementary metal-oxide-semiconductor (CMOS) integrated circuits (ICs) have obtained increased attention in the biosensor community because CMOS technology provides cost-effective and high-performance signal processing at a mass-production level. In order to supply biological samples and reagents effectively to the sensing elements, there are increasing demands for seamless integration of microfluidic components on the fabricated CMOS wafers by post-processing. Although the PDMS microfluidic channels replicated from separately prepared silicon mold can be typically aligned and bonded onto the CMOS wafers, it remains challenging owing the inherently limited aligning accuracy ( > ± 10 μm) between the two layers. Here we present a new post-processing method to create three-dimensional microfluidic components using two different polarities of photoresists, an epoxy-based negative SU-8 photoresist and positive SPR220-7 photoresist. The positive photoresist serves as a sacrificial layer and the negative photoresist was utilized as a structural material to generate three-dimensional structures. Because both photoresists are patterned using a standard photolithography technology, the dimensions of the structures can be effectively controlled as well as the alignment accuracy, moreover, is dramatically improved (< ± 2 μm) and appropriately can be adopted as an alternative post-processing method. To validate the proposed processing method, we applied this technique to build cell-trapping structures. The SU8 photoresist was mainly used to generate structures and the SPR photoresist was used as a sacrificial layer to generate sub-channel in the SU8, allowing fluid to pass through. The sub-channel generated by etching the sacrificial layer works as a cell-capturing site. The well-controlled dimensions enabled single-cell capturing on each site and high-accuracy alignment made cells trapped exactly on the sensing units of CMOS biosensors.

Keywords: SU-8, microfluidic, MEMS, microfabrication

Procedia PDF Downloads 522

Authors: Eiman Kattan, Hong Wei

Abstract:

In using Convolutional Neural Network (CNN) for classification, there is a set of hyperparameters available for the configuration purpose. This study aims to evaluate the impact of a range of parameters in CNN architecture i.e. AlexNet on land cover classification based on four remotely sensed datasets. The evaluation tests the influence of a set of hyperparameters on the classification performance. The parameters concerned are epoch values, batch size, and convolutional filter size against input image size. Thus, a set of experiments were conducted to specify the effectiveness of the selected parameters using two implementing approaches, named pertained and fine-tuned. We first explore the number of epochs under several selected batch size values (32, 64, 128 and 200). The impact of kernel size of convolutional filters (1, 3, 5, 7, 10, 15, 20, 25 and 30) was evaluated against the image size under testing (64, 96, 128, 180 and 224), which gave us insight of the relationship between the size of convolutional filters and image size. To generalise the validation, four remote sensing datasets, AID, RSD, UCMerced and RSCCN, which have different land covers and are publicly available, were used in the experiments. These datasets have a wide diversity of input data, such as number of classes, amount of labelled data, and texture patterns. A specifically designed interactive deep learning GPU training platform for image classification (Nvidia Digit) was employed in the experiments. It has shown efficiency in both training and testing. The results have shown that increasing the number of epochs leads to a higher accuracy rate, as expected. However, the convergence state is highly related to datasets. For the batch size evaluation, it has shown that a larger batch size slightly decreases the classification accuracy compared to a small batch size. For example, selecting the value 32 as the batch size on the RSCCN dataset achieves the accuracy rate of 90.34 % at the 11th epoch while decreasing the epoch value to one makes the accuracy rate drop to 74%. On the other extreme, setting an increased value of batch size to 200 decreases the accuracy rate at the 11th epoch is 86.5%, and 63% when using one epoch only. On the other hand, selecting the kernel size is loosely related to data set. From a practical point of view, the filter size 20 produces 70.4286%. The last performed image size experiment shows a dependency in the accuracy improvement. However, an expensive performance gain had been noticed. The represented conclusion opens the opportunities toward a better classification performance in various applications such as planetary remote sensing.

Keywords: CNNs, hyperparamters, remote sensing, land cover, land use

Procedia PDF Downloads 168

Authors: Francis Ojo Ologunagba, Monday Olatunbosun Ale, Lewis A. Olutayo

Abstract:

The economic successes of commercial development of agricultural product processing depend upon the adaptability of each processing stage to mechanization. In maize processing, one of its post-harvest operations that is still facing a major challenge is dehusking. Therefore, a maize dehusking machine that could replace the prevalent traditional method of dehusking maize in developing countries, especially Nigeria was designed, constructed and tested at the Department of Agricultural and Bio-Environmental Engineering Technology, Rufus Giwa Polytechnic, Owo. The basic features of the machine are feeding unit (hopper), housing frame, dehusking unit, drive mechanism and discharge outlets. The machine was tested with maize of 50mm average diameter at 13% moisture content and 2.5mm machine roller clearance. Test results showed appreciable performance with the dehusking efficiency of 92% and throughput capacity of 200 Kg/hr at a machine speed of 400rpm. The estimated production cost of the machine at the time of construction is forty-five thousand, one hundred and eighty nairas (₦45,180) excluding the cost of the electric motor. It is therefore recommended for small and medium-scale maize farmers and processors in Nigeria.

Keywords: construction, dehusking, design, efficiency, maize

Procedia PDF Downloads 323

Authors: Mohd Iz'aan Paiz Bin Zamri, Anis Salwa Mohd Khairuddin, Norrima Mokhtar, Rubiyah Yusof

Abstract:

Effective statistical feature extraction and classification are important in image-based automatic inspection and analysis. An automatic wood species recognition system is designed to perform wood inspection at custom checkpoints to avoid mislabeling of timber which will results to loss of income to the timber industry. The system focuses on analyzing the statistical pores properties of the wood images. This paper proposed a fuzzy-based feature extractor which mimics the experts’ knowledge on wood texture to extract the properties of pores distribution from the wood surface texture. The proposed feature extractor consists of two steps namely pores extraction and fuzzy pores management. The total number of statistical features extracted from each wood image is 38 features. Then, a backpropagation neural network is used to classify the wood species based on the statistical features. A comprehensive set of experiments on a database composed of 5200 macroscopic images from 52 tropical wood species was used to evaluate the performance of the proposed feature extractor. The advantage of the proposed feature extraction technique is that it mimics the experts’ interpretation on wood texture which allows human involvement when analyzing the wood texture. Experimental results show the efficiency of the proposed method.

Keywords: classification, feature extraction, fuzzy, inspection system, image analysis, macroscopic images

Procedia PDF Downloads 425

Authors: Atilla Tekin

Abstract:

Amorium was one of the biggest cities of Byzantine Empire, located under and around the modern village of Hisarköy, Emirdağ, Afyonkarahisar Province, Turkey. It was situated on the routes of trades and Byzantine military road from Constantinople to Cilicia. In addition, it was on the routes of trades and a center of bishopric. After Arab invasion, Amorium gradually lost importance. The research consists of 1372 pieces of glass bracelet finds from mostly at 1998- 2009 excavations. Most of them were found as glass bracelets fragments. The fragments are of various size, forms, colors, and decorations. During the research, they were measured and grouped according to their crossings, at first. After being photographed, they were sketched by Adobe Illustrator and decoupaged by Photoshop. All forms, colors, and decorations were specified and compared to each other. Thus, they have been tried to be dated and uncovered the place of manufacture. The importance of the research is presenting the perception of image and admiration and comparing with other cultures.

Keywords: Amorium, glass bracelets, image, Byzantine empire, jewelry

Procedia PDF Downloads 196

Authors: Yin Zhang, Kai Qiao, Xiyang Zhi, Jinnan Gong, Jianming Hu

Abstract:

In order to effectively detect aerial targets over long distances, an optimal matching design method of space-based optical payload is proposed. Firstly, main factors affecting optical detectability of small targets under complex environment are analyzed based on the full link of a detection system, including band center, band width and spatial resolution. Then a performance characterization model representing the relationship between image signal-to-noise ratio (SCR) and the above influencing factors is established to describe a detection system. Finally, an optimal matching design example is demonstrated for a typical aerial target by simulating and analyzing its SCR under different scene clutter coupling with multi-scale characteristics, and the optimized detection band and spatial resolution are presented. The method can provide theoretical basis and scientific guidance for space-based detection system design, payload specification demonstration and information processing algorithm optimization.

Keywords: space-based detection, aerial targets, optical system design, detectability characterization

Procedia PDF Downloads 168

Authors: Devineni Vijay Bhaskar, Yendluri Raja

Abstract:

We recommend a model for traffic sign detection stranded on Convolutional Neural Networks (CNN). We first renovate the unique image into the gray scale image through with support vector machines, then use convolutional neural networks with fixed and learnable layers for revealing and understanding. The permanent layer can reduction the amount of attention areas to notice and crop the limits very close to the boundaries of traffic signs. The learnable coverings can rise the accuracy of detection significantly. Besides, we use bootstrap procedures to progress the accuracy and avoid overfitting problem. In the German Traffic Sign Detection Benchmark, we obtained modest results, with an area under the precision-recall curve (AUC) of 99.49% in the group “Risk”, and an AUC of 96.62% in the group “Obligatory”.

Keywords: convolutional neural network, support vector machine, detection, traffic signs, bootstrap procedures, precision-recall curve

Procedia PDF Downloads 122

Authors: Archit Yajnik

Abstract:

Part of Speech Tagging has always been a challenging task in the era of Natural Language Processing. This article presents POS tagging for Nepali text using Hidden Markov Model and Viterbi algorithm. From the Nepali text, annotated corpus training and testing data set are randomly separated. Both methods are employed on the data sets. Viterbi algorithm is found to be computationally faster and accurate as compared to HMM. The accuracy of 95.43% is achieved using Viterbi algorithm. Error analysis where the mismatches took place is elaborately discussed.

Keywords: hidden markov model, natural language processing, POS tagging, viterbi algorithm

Procedia PDF Downloads 329

Authors: Mohammad Reza Haji Babai, Sepideh Ahmadkhan Beigi

Abstract:

Internationalization and modernization of the globe have played their roles in the process of cultural interaction between globalized societies and, consequently, found their way to the world of literature under the name of ‘imagology’. Imagology has made it possible for the reader to understand the author’s thoughts and judgments of others. The present research focuses on the intercultural images portrayed in the novel of a popular Turkish-French writer, Elif Shafak, about the lifestyle, traditions, habits, and social norms of Turkish, Americans, and Armenians. The novel seeks to articulate a more intricate multicultural memory of Turkishness by grieving over the Armenian massacre. This study finds that, as a mixture of multiple lifestyles and discourses, The Bastard of Istanbul reflects not only images of oriental culture but also occidental cultures. This means that the author has attempted to maintain selfhood through historical and cultural recollection, which resulted in constructing the self and another identity.

Keywords: imagology, Elif Shafak, The Bastard of Istanbul, self-image, other-image

Procedia PDF Downloads 141

Authors: Pooya Niksiar, Ali Ashrafizadeh, Mehrzad Shams, Amir Hossein Madani

Abstract:

There are many computationally demanding applications in science and engineering which need efficient algorithms implemented on high performance computers. Recently, Graphics Processing Units (GPUs) have drawn much attention as compared to the traditional CPU-based hardware and have opened up new improvement venues in scientific computing. One particular application area is Computational Fluid Dynamics (CFD), in which mature CPU-based codes need to be converted to GPU-based algorithms to take advantage of this new technology. In this paper, numerical solutions of two classes of discrete fluid flow models via both CPU and GPU are discussed and compared. Test problems include an Eulerian model of a two-dimensional incompressible laminar flow case and a Lagrangian model of a two phase flow field. The CUDA programming standard is used to employ an NVIDIA GPU with 480 cores and a C++ serial code is run on a single core Intel quad-core CPU. Up to two orders of magnitude speed up is observed on GPU for a certain range of grid resolution or particle numbers. As expected, Lagrangian formulation is better suited for parallel computations on GPU although Eulerian formulation represents significant speed up too.

Keywords: CFD, Eulerian formulation, graphics processing units, Lagrangian formulation

Procedia PDF Downloads 416

Authors: Hyohun Kim, Dongwha Shin, Yeonseok Kim, Ji-Su Ahn, Kensuke Nakamura, Dongeun Choi, Byung-Woo Hong

Abstract:

Artefacts are commonly encountered in the imaging process of clinical computed tomography (CT) where the artefact refers to any systematic discrepancy between the reconstructed observation and the true attenuation coefficient of the object. It is known that CT images are inherently more prone to artefacts due to its image formation process where a large number of independent detectors are involved, and they are assumed to yield consistent measurements. There are a number of different artefact types including noise, beam hardening, scatter, pseudo-enhancement, motion, helical, ring, and metal artefacts, which cause serious difficulties in reading images. Thus, it is desired to remove nuisance factors from the degraded image leaving the fundamental intrinsic information that can provide better interpretation of the anatomical and pathological characteristics. However, it is considered as a difficult task due to the high dimensionality and variability of data to be recovered, which naturally motivates the use of machine learning techniques. We propose an image restoration algorithm based on the deep neural network framework where the denoising auto-encoders are stacked building multiple layers. The denoising auto-encoder is a variant of a classical auto-encoder that takes an input data and maps it to a hidden representation through a deterministic mapping using a non-linear activation function. The latent representation is then mapped back into a reconstruction the size of which is the same as the size of the input data. The reconstruction error can be measured by the traditional squared error assuming the residual follows a normal distribution. In addition to the designed loss function, an effective regularization scheme using residual-driven dropout determined based on the gradient at each layer. The optimal weights are computed by the classical stochastic gradient descent algorithm combined with the back-propagation algorithm. In our algorithm, we initially decompose an input image into its intrinsic representation and the nuisance factors including artefacts based on the classical Total Variation problem that can be efficiently optimized by the convex optimization algorithm such as primal-dual method. The intrinsic forms of the input images are provided to the deep denosing auto-encoders with their original forms in the training phase. In the testing phase, a given image is first decomposed into the intrinsic form and then provided to the trained network to obtain its reconstruction. We apply our algorithm to the restoration of the corrupted CT images by the artefacts. It is shown that our algorithm improves the readability and enhances the anatomical and pathological properties of the object. The quantitative evaluation is performed in terms of the PSNR, and the qualitative evaluation provides significant improvement in reading images despite degrading artefacts. The experimental results indicate the potential of our algorithm as a prior solution to the image interpretation tasks in a variety of medical imaging applications. This work was supported by the MISP(Ministry of Science and ICT), Korea, under the National Program for Excellence in SW (20170001000011001) supervised by the IITP(Institute for Information and Communications Technology Promotion).

Keywords: auto-encoder neural network, CT image artefact, deep learning, intrinsic image representation, noise reduction, total variation

Procedia PDF Downloads 190

Authors: Tien-Ching Yao, Ching-Ching Lu

Abstract:

Understanding human language processing is one of the main challenges of current cognitive neuroscience. The aims of the present study were to use a sentence decision task combined with event-related potentials to investigate the psychological reality of "cross-sentence patterns." Therefore, we take the math word problems the experimental materials and use the ERPs' P600 component to verify. In this study, the experimental material consisted of 200 math word problems with three different conditions were used ( multiplication word problems、division word problems type 1、division word problems type 2 ). Eighteen Mandarin native speakers participated in the ERPs study (14 of whom were female). The result of the grand average waveforms suggests a later posterior positivity at around 500ms - 900ms. These findings were tested statistically using repeated measures ANOVAs at the component caused by the stimulus type of different questions. Results suggest that three conditions present significant (P < 0.05) on the Mean Amplitude, Latency, and Peak Amplitude. The result showed the characteristic timing and posterior scalp distribution of a P600 effect. We interpreted these characteristic responses as the psychological reality of "cross-sentence patterns." These results provide insights into the sentence processing issues in linguistic theory and psycholinguistic models of language processing and advance our understanding of how people make sense of information during language comprehension.

Keywords: language processing, sentence comprehension, event-related potentials, cross-sentence patterns

Procedia PDF Downloads 148

Authors: Seyed Mehdi Seyed Alizadeh, Christoph Arns, Shane Latham

Abstract:

Geomechanical characterization of rocks in detail and its possible implications on flow properties is an important aspect of reservoir characterization workflow. In order to gain more understanding of the microstructure evolution of reservoir rocks under stress a series of axisymmetric triaxial tests were performed on two different analogue rock samples. In-situ compression tests were coupled with high resolution micro-Computed Tomography to elucidate the changes in the pore/grain network of the rocks under pressurized conditions. Two outcrop sandstones were chosen in the current study representing a various cementation status of well-consolidated and weakly-consolidated granular system respectively. High resolution images were acquired while the rocks deformed in a purpose-built compression cell. A detailed analysis of the 3D images in each series of step-wise compression tests (up to the failure point) was conducted which includes the registration of the deformed specimen images with the reference pristine dry rock image. Digital Image Correlation (DIC) technique based on the intensity of the registered 3D subsets and particle tracking are utilized to map the displacement fields in each sample. The results suggest the complex architecture of the localized shear zone in well-cemented Bentheimer sandstone whereas for the weakly-consolidated Castlegate sandstone no discernible shear band could be observed even after macroscopic failure. Post-mortem imaging a sister plug from the friable rock upon undergoing continuous compression reveals signs of a shear band pattern. This suggests that for friable sandstones at small scales loading mode may affect the pattern of deformation. Prior to mechanical failure, the continuum digital image correlation approach can reasonably capture the kinematics of deformation. As failure occurs, however, discrete image correlation (i.e. particle tracking) reveals superiority in both tracking the grains as well as quantifying their kinematics (in terms of translations/rotations) with respect to any stage of compaction. An attempt was made to quantify the displacement field in compression using continuum Digital Image Correlation which is based on the reference and secondary image intensity correlation. Such approach has only been previously applied to unconsolidated granular systems under pressure. We are applying this technique to sandstones with various degrees of consolidation. Such element of novelty will set the results of this study apart from previous attempts to characterize the deformation pattern in consolidated sands.

Keywords: deformation mechanism, displacement field, shear behavior, triaxial compression, X-ray micro-CT

Procedia PDF Downloads 189

Authors: Stepan Papacek, Jiri Jablonsky, Radek Kana, Ctirad Matonoha, Stefan Kindermann

Abstract:

FRAP (Fluorescence Recovery After Photobleaching) is a widely used measurement technique to determine the mobility of fluorescent molecules within living cells. While the experimental setup and protocol for FRAP experiments are usually fixed, data processing part is still under development. In this paper, we formulate and solve the problem of data selection which enhances the processing of FRAP images. We introduce the concept of the irrelevant data set, i.e., the data which are almost not reducing the confidence interval of the estimated parameters and thus could be neglected. Based on sensitivity analysis, we both solve the problem of the optimal data space selection and we find specific conditions for optimizing an important experimental design factor, e.g., the radius of bleach spot. Finally, a theorem announcing less precision of the integrated data approach compared to the full data case is proven; i.e., we claim that the data set represented by the FRAP recovery curve lead to a larger confidence interval compared to the spatio-temporal (full) data.

Keywords: FRAP, inverse problem, parameter identification, sensitivity analysis, optimal experimental design

Procedia PDF Downloads 278

Authors: Avatharam Ganivada, Krishna Shah

Abstract:

The Mask R-CNN is recently introduced by the team of Facebook AI Research (FAIR), which is mainly concerned with instance segmentation in images. Here, the Mask R-CNN is based on ResNet and feature pyramid network (FPN), where a single dropout method is employed. This paper provides a modified Mask R-CNN by adding multiple dropout methods into the Mask R-CNN. The proposed model has also utilized the concepts of Resnet and FPN to extract stage-wise network feature maps, wherein a top-down network path having lateral connections is used to obtain semantically strong features. The proposed model produces three outputs for each object in the image: class label, bounding box coordinates, and object mask. The performance of the proposed network is evaluated in the segmentation of every instance in images using COCO and cityscape datasets. The proposed model achieves better performance than the state-of-the-networks for the datasets.

Keywords: instance segmentation, object detection, convolutional neural networks, deep learning, computer vision

Procedia PDF Downloads 73

Authors: Xulu Yao, Moi Hoon Yap, Yanlong Zhang

Abstract:

As a branch of artificial neural network, deep learning is widely used in the field of image recognition, but the lack of its dataset leads to imperfect model learning. By analysing the data scale requirements of deep learning and aiming at the application in GUI generation, it is found that the collection of GUI dataset is a time-consuming and labor-consuming project, which is difficult to meet the needs of current deep learning network. To solve this problem, this paper proposes a semi-supervised deep learning model that relies on the original small-scale datasets to produce a large number of reliable data sets. By combining the cyclic neural network with the generated countermeasure network, the cyclic neural network can learn the sequence relationship and characteristics of data, make the generated countermeasure network generate reasonable data, and then expand the Rico dataset. Relying on the network structure, the characteristics of collected data can be well analysed, and a large number of reasonable data can be generated according to these characteristics. After data processing, a reliable dataset for model training can be formed, which alleviates the problem of dataset shortage in deep learning.

Keywords: GUI, deep learning, GAN, data augmentation

Procedia PDF Downloads 184

Authors: Eugene Y. J. Aw, J. W. Koh, S. H. Chew, K. E. Chua, Lionel L. J. Ang, Algernon C. S. Hong, Danette S. E. Tan, Grace H. B. Foo, K. Q. Hong, L. M. Cheng, M. L. Leong

Abstract:

This paper presents a novel rapid soil classification technique that combines computer vision with four-probe soil electrical resistivity method and cone penetration test (CPT), to improve the accuracy and productivity of on-site classification of excavated soil. In Singapore, excavated soils from local construction projects are transported to Staging Grounds (SGs) to be reused as fill material for land reclamation. Excavated soils are mainly categorized into two groups (“Good Earth” and “Soft Clay”) based on particle size distribution (PSD) and water content (w) from soil investigation reports and on-site visual survey, such that proper treatment and usage can be exercised. However, this process is time-consuming and labour-intensive. Thus, a rapid classification method is needed at the SGs. Computer vision, four-probe soil electrical resistivity and CPT were combined into an innovative non-destructive and instantaneous classification method for this purpose. The computer vision technique comprises soil image acquisition using industrial grade camera; image processing and analysis via calculation of Grey Level Co-occurrence Matrix (GLCM) textural parameters; and decision-making using an Artificial Neural Network (ANN). Complementing the computer vision technique, the apparent electrical resistivity of soil (ρ) is measured using a set of four probes arranged in Wenner’s array. It was found from the previous study that the ANN model coupled with ρ can classify soils into “Good Earth” and “Soft Clay” in less than a minute, with an accuracy of 85% based on selected representative soil images. To further improve the technique, the soil strength is measured using a modified mini cone penetrometer, and w is measured using a set of time-domain reflectometry (TDR) probes. Laboratory proof-of-concept was conducted through a series of seven tests with three types of soils – “Good Earth”, “Soft Clay” and an even mix of the two. Validation was performed against the PSD and w of each soil type obtained from conventional laboratory tests. The results show that ρ, w and CPT measurements can be collectively analyzed to classify soils into “Good Earth” or “Soft Clay”. It is also found that these parameters can be integrated with the computer vision technique on-site to complete the rapid soil classification in less than three minutes.

Keywords: Computer vision technique, cone penetration test, electrical resistivity, rapid and non-destructive, soil classification

Procedia PDF Downloads 218