Hybrid Approach for Moving Object detection and Retrieving in Video

Object detection is the process offending moving objects such as animals, human beings and vehicles in videos. Object detection algorithms are typically used to extracted features and learning algorithms to recognize instances of the objects. The moving object detection is commonly used in applications such as security, surveillance, automated vehicle parking systems and image retrieval. The moving object detection has application in more areas of computer vision, video surveillance and image retrieval. The input video is converted into frames. The video is pre-processed to remove the noises from the frames. The moving objects are detected from the frames by initially identifying the object position to applying background subtraction. The morphological operations and thresholding operation were applied to extract the moving object alone. The object pixels are grouped and identified positions then represented in different color in order to obtain them exactly. A new approach to track the moving object in crowded scenes. The sequential labelling outperforms the state-of-the-art on challenging real-world videos of football, volleyball and pedestrians walking.