Lae-Jeong Park
Reduction of False Positives in HeadShoulder Detection Based on MultiPart Color Segmentation
2068 - 2073
2016
10
12
International Journal of Computer and Information Engineering
https://publications.waset.org/pdf/10006008
https://publications.waset.org/vol/120
World Academy of Science, Engineering and Technology
The paper presents a method that utilizes figureground color segmentation to extract effective global feature in terms of false positive reduction in the headshoulder detection. Conventional detectors that rely on local features such as HOG due to realtime operation suffer from false positives. Color cue in an input image provides salient information on a global characteristic which is necessary to alleviate the false positives of the local feature based detectors. An effective approach that uses figureground color segmentation has been presented in an effort to reduce the false positives in object detection. In this paper, an extended version of the approach is presented that adopts separate multipart foregrounds instead of a single prior foreground and performs the figureground color segmentation with each of the foregrounds. The multipart foregrounds include the parts of the headshoulder shape and additional auxiliary foregrounds being optimized by a search algorithm. A classifier is constructed with the feature that consists of a set of the multiple resulting segmentations. Experimental results show that the presented method can discriminate more false positive than the single prior shapebased classifier as well as detectors with the local features. The improvement is possible because the presented approach can reduce the false positives that have the same colors in the head and shoulder foregrounds.
Open Science Index 120, 2016