基于计算机视觉的番茄颜色检测研究

Tomato Color Detection Research Based on Computer Vision

以提高番茄人工分拣效率为目的,利用计算机视觉技术,建立番茄检测试验平台,将采集的番茄图像转化为HIS颜色模型,进行二值化,研究番茄成熟度。文章对番茄规定的四个等级进行分级,其分级的结果为:超优、优、良和未成熟的判断正确率分别为80%、60%、40%和80%,总的判断正确率为65%。

For the purpose to improve the efficiency of manual sorting tomatoes, using computer vision technology, to establish tomatoes test platform, to collect images of tomatoes into HIS color model with binarization to study the tomato maturity. Four grades as provided in this paper for tomatoes, the classification results include the super excellent, excellent, good and immature, the judgment accuracy were 80%, 60%, 40% and 80%, the overall judgment accuracy is 65%.