用遗传算法训练的多层前馈神经网络对番茄成熟度进行自动检测的研究

Automated Identification of Tomato Maturation Using Multilayer Feedforward Neural Network With Genetic Algorithms(GA)

建立了计算机视觉系统获取番茄的图像 ,并将 RGB值转换成 H IS值 ,通过 H 值提取番茄表面颜色特征。采用遗传算法训练的多层前馈神经网络实现番茄成熟度的自动判别。实验表明 :该方法准确率达到 94%。

Computer vision system was set up for tomato images. By using this system the RGB values of tomato hues were converted into HIS values. Automated identification of tomato maturation using multilayer feedforward neural network with GA can be realized. The results of experiment showed that the accuracy is 94%.