An Image Preprocessing Algorithm of Weak and Small Target in Digital Video Interpretation System

The weak and small target image preprocessing in the digital video interpretation system is studied. In order to improve the detection ratio of single image, firstly, morphological wavelet transform is adopted to enhance small target in the low frequency image; and then, use morphological open-reconstruction filter and threshold technique to suppress the background and segment the small target; finally, use wavelet reconstruction to get the small target detection image. Experimental results indicate that the method can detect small target precisely compared with other methods in image of low contrast small target with background clutter. The correct detection ratio of single image is 94%, and this method can improve the performance of video interpretation system.