Target Automatic Tracking Technology In The Application And Research

In the future war of high technology, people on the armored vehicle must process abundant information in the moment of battling against the enemy target, so an automatic target tracking system is needed which hangs sensors, processor and display and so on together. Such a system should extract the target from complicated and chaotic background quickly and reliably, thereby people on vehicle can react to the target as quickly as possible.Automatic tracking is a kind of comprehensive technology which consists of advanced research results in area of image processing, pattern recognition, artificial intelligence, automatic control and VLSI, etc.. Automatic tracking is the key technology for weapon intelligentization and battle automation. Based on the analysis of object’s moving images, object automatic tracking system detects the information of object’s position and moving parameters. Based on the information, system elicits the error signals between lines of tracking and sight, and controls the line of sight, thereby the line of sight could tracks the object automatically.Based on the detailed analysis of current tracking algorithms, a tracking method which integrates correlation tracking and frame differencing is proposed. The key techniques and difficulties such as correlation algorithm’s improvement for real-time requirement and template updating strategy are discussed and resolved. Function simulation is performed by Visual C++to verify the proposed tracking algorithm.In this dissertation, the digital image processing technology is explained in detail, and the working principle and structural characteristic of an image processor are described. A systematic design of the real-time image processor based on TI’s TMS320C6201is presented according to the requirements of practical project. The performance of the real-time image processor was proved by the results of practical tests.