The Image Segmentation Algorithm of Region Growing and Wavelet Transform Modulus Maximum

Compared with visible light image, infrared image eliminates the effect of the color, texture, and illumination. At the same time, the target´s brightness in general and background have more obvious differences. So a method of infrared image segmentation based on wavelet transform modulus maxima and region growing is proposed. The traditional seed region growing method regards the image pixels as the seeds to grow. And to be different with it, the method obtains the growing point and growth threshold by calculating. Firstly the method applies the wavelet modulus maximum algorithm to carry on image segmentation initialization, and forms consequent of segmentation. Then the method uses the image pixels as growing points which the gray value is greater than the average gray value of image edge point after segmentation, and uses the mean square error as growing conditions threshold which the mean square error is the average gray value between the original image and each edge point to conduct region growing. The segmentation that its characteristics combine the Wavelet Transform modulus maximum and region growing can get a better segmentation effect.