Fast Two-Dimensional Otsu's Thresholding Method Based on Integral Image

摘要

The traditional two-dimensional Otsu's threshold method using gray level-average gray level two-dimensional histogram,in view of the obvious deficiencies of the regional division of the traditional two-dimensional histogram, an improved fast two-dimensional Otsu's thresholding method with integral image based on gray level-gradient two-dimensional histogram is proposed. Using integral image to reduce the computational complexity in searching the best threshold value of two-dimensional histogram, thereby reducing the amount of calculation, the results show that the method has a good segmentation effect which greatly and significantly increased the computing speed. Therefore, it is a fast and effective thresholding segmentation method with a good real-time quality.