基于像素点对误差累加的最佳阈值选取与边界检测新算法

Best threshold value of image segmentation based on the error addition of corresponding pixels and a new arithmetic of character borderline detecting

研究存在背景噪声和干扰的情况下图像的有效分割及边界检测的问题.提出了新的图像分割准则函数——基于像素点对的误差累加最小准则和梯度二值化算法.算法兼顾了分割效果和处理的实时性,通过编程在实际图像中的应用,证明了算法的有效性和优越性。

The problem of character borderline detecting and image segmentation in an image that also contains clutter and background noise is addressed in this paper.Two arithmetics of calculating the best threshold value of image segmentation based on the error addition of corresponding pixels and detecting character borderline of the image based on the grads binarization algorithm are put forward.And the character of real time of the arithmetics is better.At last,the effectiveness of the techniques is demonstrated by applying the resulting algorithm to real world images.