基于二维最大熵和顺序滤波的红外图像分割

红外图像分割在红外成像制导和其它红外目标识别领域非常重要和提出了一种综合运用二维最大熵和顺序滤波方法对红外图像进行分割的新方法.该方法首先对图像进行基于二维最大熵的阈值分割,然后对二值图进行顺序滤波消除虚警点,取得了非常理想的分割效果.本文提出了一种新的逐步逼近二维最大熵阈值递推搜索算法,阈值搜索时间得到了大幅度减少,满足了一般的实时需求,具有较高的应用价值.

Infrared Image Fast Segmentation Based on Two-dimensional Maximum Entropy Principle and Order Filtering Algorithm

Infrared image segmentation is very crucial and important in infrared imaging guidance and other fields referring to infrared target recognition. A new infrared image segmentation method based on two-dimensional maximum entropy principle and order filtering algorithm is proposed. The method carries out the segmentation using two-dimensional maximum entropy principle at first, and then in order to remove the fault alarming point, the bitmap is order-filtered and a good segmentation result is achieved at last. Meanwhile, a new stepwise approaching and recursive threshold searching algorithm based on two-dimensional maximum entropy principle is put forward. The threshold searching time is reduced dramatically. It can meet. the common real-time demands and has high value in applications.