基于改进GAC模型的柑橘缺陷图像分割方法研究

柑橘表面缺陷的图像分割与识别是水果在收购与销售过程中的重要环节，对于提升水果的品质和经济效益有着重要意义。经典GAC模型算法虽然能够很好的分割平面缺陷，但无法准确的分割图像凹陷区域。以柑橘的虫伤、腐烂、炭疽、蓟马、溃疡5种常见缺陷图像作为研究对象，其中腐烂、炭疽为凹陷缺陷，对传统GAC模型算法进行理论研究并针对其不足提出改进方案。通过对比试验分析各参数对于试验结果的影响，并应用改进后的GAC模型算法对柑橘缺陷样本进行图像分割，分析改进后的GAC模型对柑橘表面5种缺陷的识别能力，验证改进GAC模型的可行性。

Segmentation of Citrus Defect Images Based on Improved GAC Model

Image segmentation and recognition of citrus surface defects is an important link in the marketing process of the fruit,it is also of significance to improve the quality of the fruit to enhance the added economic benefits.In this paper，five common defect images were taken as the research objects,including decay，anthrax，wounds，thrips，and ulcers caused by excessive use of drugs.Traditional model algorithm of GAC was examined,and suggestions for the improvement of the model were given.The influence of each parameter on the experimental results was analyzed by comparative experiment.The improved GAC model algorithm was applied to cut up the citrus’ defect sample in order to analyze the improved GAC model’s ability to identify five defects on the surface of citrus fruit.A series of experimental results and data verify the feasibility of improving the GAC model.