Object Plane Digital Image Processing In Tank Target Tracing And Detection

Conventional matched filter based correlator needs to make matched filter in advance, which has many disadvantages, such as large calculation and slow speed. The more important thing is that it is not suitable for real time processing and recognition precision is also very low. So the method of optical joint transform correlation is proposed, which can recognize targets automatically and is suitable for real time processing. At the same time, it can achieve higher precision.However, the conventional optical correlator is rather sensitive to the extent of complexity of the target image and the condition of its brightness. Therefore, the research of the method of image enhancement in real time joint transform correlator is always an important direction in the field.In the paper, the whole configuration of the real time joint transform correlator is described, the correlative theory of joint Fourier transform is introduced in detail, the capability of edge detection using Canny algorithm is analyzed and evaluated,and two effective methods of image enhancement are given in real time joint transform correlator. In order to avoid the defects of Canny algorithm, in the first method ,we propose an new way to remove shot noises in image by self—adaptable smoothness; in the second method ,by using variance of gray—scale to auto-adjust edge detection threshold, the image edge can be automatically detected.