色差直方图特征改进 Otsu 的橡胶缺素叶片图像分割研究

    Image segmentation of rubber deficiency leaves based on Otsu algorithm improved with characteristic of RGB dispersion

    • 摘要: 针对橡胶叶片颜色随缺素程度不断变化的特点,提出色差直方图特征改进最大类间方差法(Otsu)的橡胶缺素叶片图像分割方法。 首先利用图像 RGB 三分量计算色差图像, 保留更多的图像颜色空间信息; 然后计算图像的色差直方图, 并改进 Otsu 阈值分割算法; 最后利用改进算法对缺素叶片图像进行分割。结果表明, 改进算法在缺素叶片图像分割上优于传统 Otsu 算法。

       

      Abstract: With the degree of nutrient deficiency changing, the rubber leaf color changes. Therefore, the Otsu algorithm improved with characteristic of RGB dispersion for rubber deficiency leaves was proposed. Firstly, the image of RGB dispersion was calculated with the RGB three-component of color image, which retained more image color space information, then calculated the color histogram of image, and which was used to improve the Otsu segmentation algorithm. Finally, the deficiency leaf images were segmented by the improved algorithm. The results and algorithm analysis showed that the improved algorithm was superior to the Otsu algorithm in the image segmentation of rubber deficiency leaves.

       

    /

    返回文章
    返回