Weak Target Detection In Video Images

The dim target detection and tracking technology in video image sequence is a very important and useful. The image background clutters estimation and removing technology is a significant part of dim target detection and tracking technology. I mainly study this in this paper.What is a dim target? What are the dim target specialties? What are difficulties in dim target detection? All of these are need to be answered. There are discussions about the definition of dim target, dim target's specialties and relationship between dim target and other parts of image.Because of difficulties of dim target detection and identifying, and there are background clutters and noise interfering, the dim target is easy to be masked. So we can deal with the original image sequence, estimating the background clutters and removing them from image sequence. The realities tell us the step is needed and efficient. There are two kinds of background clutters estimation and removing methods: one is in frequency domain and another is in space domain. I discuss two methods in this paper, and think out three criteria to evaluate the different methods.This article firstly uses Wiener filter for original images, and then introduces an background estimate method based on the relativity between image sequences for filtered images to pick up the background and finally detect dynamic small targets. I simulate all algorithms by computer, and comparing their performances on clutters removing.