Real-time pedestrian detection using support vector machines

摘要: In this paper, we present a real-time pedestrian detection method in outdoor environments. It is necessary for to implement obstacle and face which are major parts of walking guidance system the visually impaired. detects foreground objects on ground, discriminates pedestrians from other noninterest objects, extracts candidate regions recognition. For effective detection, have developed using stereo-based segmentation SVM (Support Vector Machines), works well particularly binary classification problem (e.g. object detection). We used vertical edge features extracted arms, legs torso. our experiments, test results large number scenes demonstrated effectiveness proposed method.