Abstract:

The infrared measurement image is usually a gray image, and pseudo color processing is a very important means of enhancement. The traditional pseudo color processing methods do not consider grayscale breakpoints between the high temperature target and low temperature background in the actual image, after pseudo color processing, the number of colors of the image is always small, and lots of details information is lost. In order to get better display effects of the high temperature target, its edge details and the low temperature background, according to the gray level, the image is divided into the target area, background area and targetbackground transition area, and the background area contains multiple transition areas between the backgrounds. The K-means clustering theory is used to cluster the background. According to the clustering results, confirm image gray level ranges of the multiple background area and transition area by the gray level compression method. According to the statistical characteristics of each interval gray, map the multistage linear gray level for the whole image, and balance the image grayscale distribution effectively. Introducing visual difference factors, use the CIE94 color difference formula modify the traditional color path, realize pseudo color display of the image in equal color difference. Experimental results show that, the method is more suitable for the high temperature target infrared measurement image than the traditional method, the target of the processed image is outstanding, the color of image details is abundant, and the contrast of background is high, it can largely enhance the ability of recognition with infrared measurement image details for high temperature targets.