Abstract：To solve the disadvantages of traditional Otsu algorithm with large computation and bad realtime performance, a new PSO+Otsu(S) segmentation method was proposed. The RGB color space was converted into HSV color space, and the S component image was extracted. A new updating strategy of inertia weight coefficient w was proposed for the basic particle swarm optimization algorithm, and the improved particle swarm optimization algorithm was used to search the optimal threshold. The optimal threshold was used as the threshold of Otsu algorithm, and the Otsu algorithm was used to segment the S component image. Finally, the segmentation image of lettuce leaf was obtained. The results show that the proposed segmentation method is not only suitable for single leaf image, but also suitable for canopy leaf image. When the algorithm is used to segment the single image of lettuce leaf and the canopy image of lettuce leaf, the program running times are 118 ms and 126 ms, and the iteration numbers are 6 and 5 times, respectively. Compared with the Otsu algorithm and the standard PSO+Otsu algorithm, the proposed algorithm can not only shorten program running time, but also improve the accuracy of image segmentation with better realtime performance.
2018, Vol. 39 
