Image based on histogram and K-Means clustering segmentation algorithm

The traditional K-Means algorithm is mainly used for image segmentation with large differences in color. Since the traditional K-Means clustering algorithm is easy to be sensitive to noise and it is difficult to obtain the optimal initial cluster center position and number, a method based on histogram and K-Means clustering is proposed. The median value is used. Filtering and histogram averaging preprocess the image to reduce the influence of noise on the pixel; then estimate the number of centers based on the number of peaks appearing in the histogram obtained after preprocessing, and then determine the aggregation in the Lab color space. At the center of the class, the cluster center is continuously iterated and updated until the algorithm finally converges to realize the clustering operation. The experimental results show that the improved algorithm shortens the time of the clustering process and can effectively extract the target object and improve the image segmentation effect.