基于改进OTSU的煤尘颗粒图像分割算法研究

Research on Image Segmentation Algorithm of Coal Dust Particles Based on Improved OTSU

针对传统的最大类间方差法(OTSU)进行图像处理时会出现部分煤尘颗粒不显示的问题。论文对最大类间方差法(OTSU)进行改进,先利用改进的同态滤波去除图像噪声和光照不均匀情况,增强图像对比度,再通过改进的最大类间方差法对图像进行分割处理。论文提出的方法所得到的图像对比度为358.47450,比传统同态滤波方法提高45.76285,结果更优。论文提出的算法可以使整幅图像的煤尘颗粒显示的更完全、更具体。实验证明,论文方法具有重要的理论意义和应用价值。

Some coal dust particles are not displayed when image processing is performed on the traditional maximum between-class variance method(OTSU).In this paper,the maximum between-class variance method(OTSU)is improved.First,the improved homomorphic filter is used to remove image noise and uneven illumination,and the image contrast is enhanced,and then the image is segmented by the improved maximum between-class variance method.The image contrast obtained by the method proposed in this paper is 358.47450,which is 45.76285 higher than the traditional homomorphic filtering method,and the result is better.The algorithm proposed in this paper can make the whole image of coal dust particles displayed more completely and concretely.Experiments prove that the method in this paper has important theoretical significance and application value.