Detection and Retrieval of Colored Object from a Live Video Stream with Mutual Information

Abstract

Detection and segmentation refer to a process of segregating a part of any body from its whole. In the case of image segmentation, it refers to the method of obtaining a particular portion of that image. As we know, a video or any mp4 file is an orderly orientation of frames passing the screen per second; hence, video detection schemes can also use this segmentation process. With the advancements of AI and machine learning in our daily life, video segmentation has become a necessity for much technological advancement. Some of the important applications of video segmentation could be a face recognition system, color detection schemes, an object tracking system, etc. If we want to track anyone or some object, video surveillance becomes the primary key to such an application, and hence, a lot of importance is given to it. To find the best results on object detection techniques, different algorithms put forward. Here we would cover all the technological advancements and researches, works done on object detection and live stream of videos. With today’s computing powers, the vision to tracking objects as well as moving objects has advanced a lot. Various needs like video surveillance and facial recognition systems in many security checks use color detection and object recognition schemes. Although there are many challenges like high resolution of the videos, machines capable of analyzing higher frame rates in the videos and many more.