Air And Sea Target Recognition Algorithm Research And Directshow Achieve

Target identification has been an ascending technical field for a long time and still be active, which is demonstrating its enormous application in the information industries. The target identification algorithm under the sea-air background is researched by this article. Because of water vapor in the air, the background light, and other factors, not only the quality of images is poor but also the objective information is unobvious. Therefore, researching the target identification technology in the sea and air conditions is very important for getting target information with high quality.A mature technology, Microsoft's DirectShow, is adapted to implement software handling sea-air targets identification in video files by this article. First of all, DirectShow technology and the basic framework of overall situation are introduced in this paper. Secondly, the original video is pretreated by image processing algorithms to reduce the noise caused by the sea-air environment. It benefits the following target detection. Thirdly, an algorithm based on the row and column average color value is used to detect a rough area with target. Then the partition area containing target would be handled by step-edge detection algorithm to accurately identify the target. This algorithm with simple logic run faster and could support real-time video processing. The result of this algorithm is satisfied. Then, one universal image processing software is programmed to simulate this algorithm based on the row and column average color value with C#. In addition, other algorithms in this article are also simulated in this software. Finally, according to the simulated algorithm, program DirectShow Filter and Filter Graph with DirectShow software develop kit to process the video file directly.