沈阳农业大学学报2016,Vol.47Issue(3):334-341,8.DOI:10.3969/j.issn.1000-1700.2016.03.012
自然环境下杨梅果实图像的分割方法研究
Research on Bayberry Image Segmentation Method in Natural Environment
摘要
Abstract
In order to solve the problem of accurate segmentation of bayberry image in natural environment, a segmentation method of bayberry fruit based on homomorphism filtering was put forwarded. Firstly, through the analysis of the color features of bayberry fruit, the R-G color different component was used to handle the bayberry fruit image based on the RGB color space. For binary image processing, the morphology and connected region labeling method were used to eliminate the possible image noise, thereby the leaves, and most background were removed. Secondly, the ratio of the length and width of the minimum bounding rectangle in the bayberry region was calculated to determine the branches in the region, for determining the need for further segmentation. Before the next segmentation step, in order to eliminate the impact of uneven illumination, image enhancement using homomorphism filtering method was used to enhance the brightness of V components in the HSV color space, and achieved illumination compensation. Finally, the R-G color components were used to segment the remaining branches and other parts of background, with the Otsu segmentation method in RGB color space and to achieve bayberry image recognition. The experimental results showed that this algorithm could identify bayberry from the background effectively, with mean segmentation error A f of 2.26%, compared with the mean segmentation error A f of 25.23% in the image segmentation algorithm, using direct Otsu segmentation method and compared with the mean segmentation error A f of 18.12% in the image segmentation algorithm, using direct homomorphism filtering, thus the effectiveness of this algorithm was verified.关键词
杨梅/R-G/色差分量/同态滤波/OtsuKey words
bayberry/R-G/color different component/homomorphism filtering/Otsu分类
信息技术与安全科学引用本文复制引用
徐黎明,吕继东..自然环境下杨梅果实图像的分割方法研究[J].沈阳农业大学学报,2016,47(3):334-341,8.基金项目
江苏省自然科学青年基金项目(BK20140266);江苏省高校自然科学研究面上项目(14KJB210001);江苏省高等职业院校国内高级访问学者计划资助项目(2014FX031);常州大学科研启动项目(ZMF13020019) (BK20140266)