中文摘要：

为了提高金刚石锯片磨粒识别的效率和质量,将BP神经网络算法应用于金刚石锯片磨粒的识别。以三维形貌的高度、梯度、二阶差分作为神经网络的输入,磨粒、结合剂作为神经网络模式识别的二类输出,应用处理过的形貌对网络进行训练,建立了磨粒识别的神经网络模型。实现了对三维金刚石磨粒的自动识别,避免了人为判断的主观性。实例表明,该方法可以有效地识别磨粒,具有较好的推广价值。

英文摘要：

In order to improve the efficiency and quality of diamond saw blade abrasive grain of recognition,BP neural network algorithm was applied to the recognition of diamond saw blade abrasive grain.Abrasive grain identification of neural network model is established through neural network input with the height,gradient and the second order difference of 3Dmorphology,second neural network output with abrasive grain and bonding agent,and training neural network with processed morphology.Thereby,it achieves automatic identification of the 3Ddiamond abrasive grain,avoids subjectivity of human judgment.Examples show that the method can efficiently identify the abrasive grain,which owns a good promotion ability.