基于分形的快速最大熵红外图像特征检测

Fast maximum entropy for characteristics detection of infrared image based on fractal theory

文章针对红外图像目标检测问题,提出一种基于分形的快速最大熵的红外图像特征检测算法.该算法利用DBC方法计算分维数,根据人造物和自然背景分形维差异,确定目标区域;最后,通过二维最大熵原则确定最佳阈值,实现对单目标或者多目标图像分割.该算法能够较好实现红外图像特征检测,有效抑制背景和噪声.

For the problem of target detection in infrared image,a fast maximum entropy for characteristics detection of infrared image based on fractal theory was proposed.The algorithm uses DBC methods to calculate fractal dimension,gets target areas based on artifacts and natural background difference fractal dimension.Finally,it determines the best threshold for single target or a target image segmentation by two dimension maximum entropy principle.The algorithm can better achieve infrared image feature detection,and effectively suppress background and noise.