The Research Of Detection Algorithm For Dim Targets In Infrared Images

Making use of infrared image to realize the automatic target examination, recognition and tracking is the main development direction in equipments of modern military weapons. Because infrared sensor is easily affected by atmosphere hot radiation, long distance and sensor noise, the detected targets in infrared images often present like dim targets, even like point targets, and drowned in noise. The dim targets detection in infrared images is one of the key techniques in military weapon system. On the other hand, dim targets detection in infrared image is also widely used in public, such as the astronomy prognosticates, particle collision, forest early warning and remote sensing etc. So the dim targets detection in infrared image has big practical value. This paper has a systemic research on the common algorithm of pre-processing and targets detection, summarized out some useful results. And considering the IR target's characteristics such as low signal-to-noise ratio, and factor that the targets near the horizon are disturbed by the background edge clutters, an improved method is presented to solve the problems in this paper. At first, noise is reduced by median filtering. Secondly, the DQS (Difference Quadratic Sum) algorithm is presented to restrain the background noise and enhance the targets. Thirdly, the horizon is filtered by line scanning and the background edge is suppressed. Then, after image threshold segmentation, the suspicious targets are extracted. Finally, the emulation experimental result is gained by programming and hardware debugging on the DAM6416P platform. And the result shows that the dim targets detection algorithm achieves a high availability and reliability in dim targets detection. At present, these researches in this paper achieves anticipative objects, and are the basis of developing the high performance targets extraction system in the future.