DETECTION AND TRACKING OF HUMANS BASED ON IMPROVED HISTOGRAM OF ORIENTED GRADIENTS AND KALMAN FILTER

Abstract

Human detection and tracking in a video surveillance system is critical for various application areas including suspicious event detection and human activity recognition. In the current environment of our society suspicious event detection is a burning issue. For that reason, this paper proposes a framework for detection and tracking of humans by generating a human feature vector. Initially, every pixel of a frame is represented as an incorporation of several Gaussians and use a probabilistic method to refurbish the representation. These Gaussian representations are then estimated to classify the background pixels from foreground pixels. Shadow regions are eliminated from foreground by utilizing a HueIntensity disparity value between background and current frame. Partial occlusion handling is utilized by color correlogram to label objects within a group. After that, the framework generates regions of interest (ROIs) by considering conditions related to human body. Afterward, features are extracted from ROI for classification. A feature descriptor, Improved Histogram of Oriented Gradients (ImHOG) is proposed to alleviate the limitation of Histogram of Oriented Gradients (HOG). Finally, Kalman filter is utilized for human tracking to increase detection rate. Various videos containing moving humans are utilized to evaluate the proposed framework and presented outcomes demonstrate the adequacy.