An Improved Video Object Segmentation and Tracking based on Features using Threshold Filtering Technique

Abstract

Video object segmentation and tracking is significant research topic in a video surveillance application. Recently, many researches has been developed for video object segmentation and detection, however, the video object segmentation based on features like shape, texture, intensity was not efficiently performed. In this paper, an Improved Threshold Filtered Video Object Detection and Tracking (ITFVODT) framework is designed for efficient video object segmentation based on their features like shape, texture, intensity and tracking of moving objects. ITFVODT framework initially takes video file as input. Then, ITFVODT framework segments the video frames based on shape, texture, intensity of image. After the object segmentation, filtering technique is applied for tracking the video objects. Filtering technique is used in ITFVODT framework for improving the video quality by reducing mean square error. Finally, ITFVODT framework performed the video objects detection task with help of Thresholding technique which in turn improves the video object detection accuracy. The proposed ITFVODT framework using video images obtained from Internet Archive 501(c) (3) for conducting experiment. The performance of ITFVODT framework is tested with the metrics such as object segmentation accuracy, Peak Signal to Noise Ratio, object tracking accuracy, Mean Square Error and object detection accuracy of moving video object frames. Experimental analysis shows that the ITFVODT framework is able to improve the video object segmentation accuracy by 12% and also improve video object detection accuracy by 17% when compared to the state-of-the-art works.