【摘要】 提出一种基于数字图像处理的黄河河工模型图像的河势宽度测量方法.该方法利用边缘检测和形态学相结合,对黄河模型图像进行边缘提取,然后利用像素计算法,通过比例转换实时精确地得出河势宽度.首先利用改进的最大方差比法(Maximum variance ratio)自适应地计算出Canny算法的高低梯度阈值,并在此基础上采用基于周长的形态学连通域分割法清除图像的干扰边缘,实现河势的自动识别和提取,再通过示踪粒子标定河道,计算河势边缘间像素个数,从而得到河势宽度.试验表明,该方法抗干扰性强,提取误差小,宽度测量更加简单.更多还原

【Abstract】 A method based on digital image processing is presented for the width measurement of Yellow river model regime.This method uses the adaptive threshold Canny edge detection algorithm and the morphological connected domain segmentation method to extract the edges of the Yellow River model image.Then,by means of the pixel calculation method and the proportion of conversion,we can real-time access to the precise width of the river regime.It firstly uses the improved maximum variance ratio method to calculate the values of Canny gradient thresholds self-adaptively.And on this basis,the algorithm of morphological connected domain segmentation based on the circumference is used to remove the interference edges of the image.This method achieves the automatic identification and extraction of model regime.And then,the number of pixels between the edge of river is calculated by the method of tracer particle calibration of channel to get the width of river regime.Reserch results showed that the algorithm had good anti-jamming performance and high extraction precision,and that width measurement would be easier.更多还原