论文部分内容阅读
为了更好的监测桥梁裂痕,文章提出了一种改进的细胞神经网络桥梁裂痕图像识别方法。该方法通过一定的图像处理,建立裂痕网络和细节网络,同时增加了裂痕相似网络模型,避免了仅对裂痕特征提取信息不能准确识别裂痕的问题。实验证明,改进的裂痕识别算法实现简单,识别桥梁裂痕准确率高,达到了实时识别技术的要求。
In order to better monitor the bridge cracks, the paper presents an improved method for bridge crack image recognition based on cellular neural networks. The method establishes a network of cracks and a detail network through certain image processing, and adds a similar network model of the cracks to avoid the problem that the cracks can not be accurately identified only by extracting the information of the cracks. Experiments show that the improved algorithm for crack identification is simple, and the accuracy of identification of bridge cracks is high, which meets the requirements of real-time identification technology.