Moving Object Detection and Counting Using Fuzzy Color Histogram Features

Object detection and counting from video stream is very important for many real-life applications. Existing detection and counting were based on Bayesian regression. We present a efficient object detection and counting based on background subtraction using fuzzy color histogram (FCH), which used effectively for removal of unwanted pixel from the background and capacity of extraordinarily weakening shade varieties created by foundation movements while as of now highlighting moving articles for effective individuals tallying. First, video is converted into frames for processing it to still images to detect objects. Fuzzy C means (FCM) technique used for data grouping, applying along with membership values for clustering with color planes [1]. Foreground and background classified with FCH features by applying threshold value ranges from 0 to 1. Then detected object proceed with morphological process and component analysis for smoothing. Finally, object is counted and we present number of objects in full video.