二代小波变换的抗噪Otsu图像分割方法

Image Segmentation of Anti-noise Otsu Based on the Second Generation Wavelet Transformation

针对一维、二维最大类间方差(Otsu)方法分割含噪声图像时分割效果不佳、抗噪性不足的问题,提出了一种二代小波变换的抗噪Otsu图像分割方法.算法通过二代小波变换,在小波域中对目标与背景的噪声进行抑制后再计算最大类间方差进行了研究.实验结果表明该算法是一种抗噪性强、分割效果好的图像分割算法.与现有的二维、三维Otsu法相比,算法不仅计算效率高,而且抗噪性能更稳健.

To solve the problem of the poor effect and anti-noise deficiency in the segmentation of noisy image with one dimension and two dimension Otsu method, this paper proposes an Otsu image segmentation method for noise reduction based on second generation wavelet transformation. The second generation wavelet transformation is used to suppress the noise of the target and background in the wavelet domain, then the otsu is calculated. The experiments show that the algorithm is a good image segmentation algorithm with strong anti-noise and good segmentation effect. Compared with the existing two-dimensional and three-dimensional Otsu methods, the algorithm is not only computationally efficient, but also robust to noise suppression.