【摘要】 提出了基于改进BP神经网络的复杂背景下的回转体目标识别方法,实现了对目标的准确提取。采用中值滤波滤除图像噪声,用改进最大类间方差阈值法进行图像分割。提取回转体目标7个不变矩特征作为神经网络输入的特征向量,建立了基于BP神经网络的目标识别系统,进行回转体目标识别,模拟实验结果说明,所提出的图像预处理方法可有效去除复杂背景图像噪声、准确地分割图像,选择7个不变矩特征作为回转体目标识别特征是合理的,基于BP神经网络的回转体目标识别方法具有较高的识别率。更多还原

【Abstract】 Rotary object recognition in complex background based on improved BP neural network is proposed in the dissertation.Median filter is adopted to get ride of the noise and an improved method of maximum classes square error is used to compute the threshold of the image segmentation.The target recognition system based on the improved BP neural network is established to recognize the rotary objects,and seven invariant moments of rotary objects serve as the input feature vector.The experimental results show that the image noise could be removed effectively and the image could be segmented exactly by the image preprocessing method put forward in the dissertation,and the seven invariant moments are appropriate for the character of rotary objects,and the rotary object recognition system based on BP neural network acquires an excellent recognition result.更多还原