Recent research trends and methods used in moving object detection

Abstract

Object detection is one of the main challenges in the field of image processing from the beginning. Many methods had been established and proposed through the years, but none of them could give a high accuracy rate along with low computation time. Background subtraction is a key technique in object detection in security and surveillance purpose with a high accuracy rate, but it comes in a cost of high computational time. Frame differencing came with low computational time, but it cannot give us the accuracy that we want. Optical flow and temporal differencing are independent of irregularities of light features, but accuracy is not up to the mark. The main obstacles of detecting an object in a video are still present in proposed techniques till date. In this paper, we mainly describe the recent trends and works done in the field of object detection. We have discussed various important methods of object detection and pointed out their positive aspects and limitations.