Human object detection by HoG, HoB, HoC and BO features

For several decades researchers around the globe have been avidly investigating practical solutions to the enduring problem of understanding visual content within an image. We have focused our work on features extraction ofimage by four different methods and cascaded the features followed by support vector machine classification algorithm. We have used sliding window approach in it to locate the object in the. The window size is kept fixed and image is scaled down to scan all objects. The four features extracted are histogram of gradient (HOG), histogram of bar (HOB), histogram of color (HOC) and block orientation (BO).