二维最大相关准则图像阈值分割递推算法

提出了基于二维直方图的最大相关准则自动阈值图像分割算法。该算法根据图像中目标和背景分布的相关量最大来选择阈值。基于二维最大相关准则的阈值分割算法具有较强的抗噪声能力,能够实现准确分割。在实际应用中,为了加快二维最大相关准则阈值分割的计算速度,减少重复计算,推导了该算法的快速递推公式。递推算法节省了计算时间,使算法具有更强的实用价值。通过对低对比度、低信噪比的遥感红外图像进行试验,结果表明二维最大相关准则图像阈值分割算法具有良好的分割效果,其递推算法使其计算速度提高了近30倍。

Recursive Algorithm of Image Thresholding Based on Two-dimensional MCC

Automatic image thresholding based on the two-dimensional Maximum Correlation Criterion(MCC)is proposed. On the base of two- dimensional histogram,this method maximizes the correlations associated with the distributions of the background and object classes to obtain the optimal threshold.The image thresholding based on two- dimensional MCC has greater resistance capability to noise and the threshold is optimal.In realistic appliance,the recursive algorithm is introduced to elevate the computational velocity and reduce the repeat computation.The recursive algorithm shorts the computational time and realizes real- time image segmentation.The experimental results show that the proposed algorithm achieves a good segmentation quality and elevates the computational velocity near 30 times.