一种基于阈值的图像分割算法研究

为了解决图像中出现的照度不均匀现象,提出了一种基于阈值的图像分割算法。首先基于边缘阈值和非边缘阈值的计算得到初步的阈值图像,然后通过拟合区域划分、控制点的简约和阈值的外插等步骤对图像实行再次分割。实验证明此算法分割图像后得到的效果好,相比于其他典型算法具有优越性,特别是边缘处的吻合程度较高,并且使图像抵抗模糊的能力增强,使图像更加平滑而没有噪声,解决了以前利用边缘方法容易被噪声妨碍的问题。

Research on an Image Segmentation Algorithm Based on Threshold

To solve the uneven illumination of image,an image segmentation algorithm based on threshold is proposed. This algo-rithm obtains the original image by computing the edge-threshold and non-edge-threshold. Then the image is segmented again through dividing the fitting region, reducing the controlling point and extrapolation of threshold. Experimental results show that the algorithm pro- vides an excellent image segmentation effect. And its Edge-Border Coincidence is higher than other typical image segmentation algo-rithms. It offers a much stronger Anti-fuzzy ability and smoother image without noise, overcoming the weakness of previous edge-based method easily impacted by noise.