Digital Image Processing In Target Detection Applications

The development of Digital Image Processing Technology (DIPT) is increasing rapidly as the enhancement of computing speed. Combining the DIPT with aim images, the application in obtaining the aim information from corresponding images and achieving the goal detection with a high probability based on DIPT is a hot research field.Detector receiving surface imaging feature is closely related to the distance between the detector and the goals. The detection method is different when the distances vary. As the basic process of target detection "Image acquisition - Pretreatment -Background Suppression - Target detection" for clues, the application of DIPT in the detection of targets is studied detailedly in this dissertation . The main achievements of this dissertation are as follows:(1) The principle of digital image acquisition and several equipments is briefly introduced;(2) For long-range detection, the topics of correcting and restoring drop image caused by atmospheric turbulence are carried out an effective discussion. The degraded model derived from the wave structure function and refractive index spectrum including the impact of internal and external scale is used to restore a drop image including the target of plane. Simulation results show that the model concluding the influence of outer-and inner-scale of turbulence is better than that derived directly from the Kolmogorov spectrum;(3) The background suppression technique of infrared small target image is studied. According to the ups and downs background statistical characteristics of infrared images, an improved method based on SVR filtering is presented in this paper. Theoretical and experimental results show that the method for suppressing infrared background is effective. After filtering, an adaptive threshold segmentation method is presented to further deal with the infrared image. And then the binary images including less suspicious target are obtained, which provides favorable conditions for the latter image sequence detection method; (4) In the basis of the research of background suppression technique, the image sequences target detection method based on pipeline structure is detailedly studied. Accordance to the contiauity of target and the consistency of the moving contrail, the binary image processed in former is detected using the location relationship of suspicious targets between adjacent frames. The simulation shows that the method is relatively ideal.