用遗传神经网络方法进行苹果颜色自动检测的研究

Study on Automated Color Inspection of Apples Using Genetic Neural Network

摘要: 该文通过计算机视觉技术获取苹果的色度直方图并提取其表面颜色特征，采用先进的遗传算法建立了一个多层前馈神经网络系统，从而实现了苹果成熟度的自动判别。实测表明，该方法准确率较高，并具有鲁棒性、灵活性和高速度。

Abstract: The external color feature of apples was extracted from the hue histograms using computer vision technology. With the advanced Genetic Algorithms, a threelayer feed forward neural network system was established, which can realize online automated judgment for maturity degree of apples. This method was proved to be robust, flexible , highspeed as well as accurate.