Analysis and comparison of automatic image focusing algorithms in digital image processing

Optical imaging equipment is widely used in many fields such as industry, medicine and military, etc. The autofocus, as the main component affecting image clarity, still needs to be optimized and improved. A simpler passive autofocus system is selected, and a grayscale covariance matrix is introduced to use its output of inter-pixel grayscale relationship data to attenuate the influence of noise and realize the visual measurement of image quality. Next, the focus window locking is optimized, and a first-order moment window method incorporating image difference shadow is proposed, in which the Gaussian blur can enhance the smoothness of the image, and the difference shadow map is obtained according to the difference between the original image and the smoothed image, and the method keyed out the background of the image to enhance the detail portrayal of the target region, so as to locate the center of focus, and finally the window locking is realized in accordance with the proposed size. The study verifies the performance of the focusing model through simulation experiments, which show that the mean value of the accuracy of the model is the best value of 0.96 among the control models in the face of the Gaussian noise of (0.1,0.004). Therefore, this autofocus model can effectively improve the autofocus efficiency.