Segmentation of infrared image using support vector machine

Segmentation of infrared image is very important in infrared image analysis. Support vector machine (SVM) approach is considered a good candidate because of its good generalization performance, especially when the number of training samples is very small and the dimension of feature space is very high. In this paper, a segmentation algorithm based on SVM for infrared image is presented. The algorithm extracts the target from the infrared image by SVM. The image is divided into the target and background. Aiming at the best performance of infrared image, detailed analysis and comparisons are made by choosing different kernel functions and correlative parameters. Experimental results show that the presented algorithm more effectively and accurately extracts the target than the existing methods. SVM has a good application prospect for the segmentation of infrared image.