An Automatic Focusing Algorithm Based on Blur Edge Width Detection in the Image

Abstract

An automatic focusing algorithm through detecting blur edge width via wavelet transform was proposed. To improve computation speed and fit for different scenes, the center area of the image and four golden points were selected as the focusing area. Wavelet module values of the focusing area were obtained through wavelet transform. The threshold to determine blur edge was selected by 50 percent of the wavelet module maximum in all wavelet module values. According to the threshold, the binary image of the blur edge was achieved. Area and girth of the blur edge were calculated respectively, and blur edge width was computed by area and girth. The control center computed steps according to blur edge width to adjust the lens, and the focusing motor operated corresponding steps received from the control center. Therefore, an automatic focusing procedure was implemented. Results show that the automatic focusing algorithm is quick and effective.