The method of fruit image segmentation and location of green sunshine rose grape(aka.Vitis Labrusca)in orchard environment was studied in this paper.After analyzing the captured images under the conditions of smooth light,backlight and night light,the color component which can best reflect the green grape fruit is selected as the input image of the segmentation algorithm.The maximum inter-class variance method OTSU and support vector machine method SVM are used to segment fruit and background regions.The experimental results show that the recognition rate of green grapes at night is higher than that of smooth light and backlight in sunny days.Comparing the accuracy of the two algorithms,we can find that the accuracy of SVM algorithm is higher in sunny days with smooth light and backlight,while OTSU algorithm is higher at night,reaching 98.7%.
green grapeimage segmentationOTSU algorithmSVM algorithmmachine vision
国家科技期刊平台
NSTL
万方数据
