IR Small Target Real-time Detection and Tracking in Clutter Sea Background

Along with the development of Infrared imaging instrument, the detection and tracking of target are widely used in missile controlling-guidance, vision of robot, video watching, medical treatment, etc.The Infrared imaging system is a passive detection system because it doesn’t radiate out electromagnetic wave. Because of this feature, the infrared imaging system has many strong points such as secretiveness and anti-jamming. Along with the development of Infrared imaging instrument, the quality of imaging is being improved, which makes requires that detect and track small target which has long distance. So the detection and tracking of small target in infrared image is becoming the hot research point.In this dissertation, we do some research in the detection and tracking of small target in infrared image which has the background of sea surface and wave clutter. The main contribution of the dissertation is as follows:(1) Research the pretreatment of infrared image which has the background of sea surface and wave clutter. Research and realized an pretreatment of image based on multilevel filter. Our experiment validate that this method can restrain the background of sea surface and wave clutter, and improve the signal-clutter rate(SCR).(2) Introduce the detection and tracking system architecture of small target in infrared image, and a detection algorithm of small target in infrared image is studied and realized. This algorithm used not only the target information in one image, but also the motion continuity of moving target in sequence image. To reduce the interference of noise and wave clutter, our algorithm classify the dubious target based on apriori knowledge. At last, we proposed an detection of abnormal cell, which is brought out because of the damage of imaging point. This abnormal cell will interfere the detection of target because that it’s similar to real target.(3) An tracking algorithm of small target in infrared image based on Kalman filter is proposed and realized in this dissertation. The tracking of small target in the reflection of sun is studied, and a method of detection of sun reflection is proposed and realized. Based on the detection of sun reflection, we proposed a method of tracking small target in the sun reflection. Our experiment validate that our method can detect the sun reflection effectively, and maintain the tracking when the sun reflection exist.