基于不变矩特征及BP神经网络的图像模式识别

Pattern Recognition Based on Moment Invariant Feature and BP Neural Network for Image

根据图像的不变矩特征,应用BP神经网络实现了图像的模式识别.由于神经网络本身具有很强的学习能力及容错能力,且采用并行工作方式,因此,此识别方法与传统的模式识别方法相比,具有较强的抗干扰能力及较快的识别速度.本文通过实验验证了此方法的有效性.

A image pattern recognition method based on moment invariant feature and BP neural networks is archived. Because of the strong learn ability, tolerance errors ability and parallel task, pattern recognition based on neural networks have strong anti-disturbance ability and fast speed. Experiment result demonstrated that the method is effective.