Research And Implementation Of The Detonation Point Of Automatic Identification System Based On Video Image

Because there are many disadvantages in present manual judgment and count fordetonation, such as tiring for operator, easy to false and missing report, this project hasfinished automatic judgment and count for detonation by real-time collecting, analyzingand processing the video of impact area.In this project, cameras were fixed to finish continuous shooting the impact area,and the video was transmitted to the command center, in which the video was extractedto continuous images. Then the automatic identifying software will process theseimages, pick up and identify the targets, display and record the identifying information.The main works include the following four aspects: First, the real-time collection andpre-processing of the video images was studied. In this step, digital filter andintersecting frame were used to get gray images for the followed process. Second,automatic exacting the faint targets against complex background was accomplished. Itwas finished by difference, filter and segmentation of gray images. Third, thetechnology of automatic identifying targets by tracks correlation was researched. At last,this project will automatic report and record the number, time and site of explosions.This project has been verified by real tests. It can release operator from hard work.With explicit judging result and high correctly identifying rate, this project will provideessential reference for the fast evaluation of test quality, which will play an importantrole in promoting the test quality.