基于区域生长的红外图像分割

Infrared Image Segmentation Based on Region Growing

该文采用区域生长的方法 ,结合人工智能 ,针对红外图像中目标相对较小、灰度值与视场中目标物体的热辐射量相关的特点 ,选取相应种子点根据灰度相似性判决生长出目标区域。结果表明 ,采用这种方法具有一定的针对性 。

A method that incorporates region growing and artificial intelligence techniques is presented.According to the peculiarity of infrared images,a pixel is chosen that has the highest gray level as the seed of growing process.The gray level decision criterion is used to perform the growing process.An experiment is made on a field infrared image of a tank.The result has shown that the method is effective to the segmentation.