基于图像处理的水位信息自动提取技术

摘要

为满足水尺量测水位自动化和实时性的需求，提出通过图像处理技术实现对水尺图像自动提取水位信息的技术方法。首先通过对水尺图像进行图像增强、二值化、边缘检测和去除噪声等处理定位出水尺；然后根据水尺上字符的特征实现字符分割，采用模板匹配法实现字符识别，并运用最长等差数列法对识别结果进行优化校正；最后根据识别结果，分不同情况计算出水尺读数。实验结果表明，该方法识别信息的正确性和精确性较高，可满足实际应用需求。%In order to meet requirements of automation and real-time of water gauge for measuring water levels, the article proposes research and technical methods of automatically extract water gauge water level image information by image processing technology . First it locates the water gauge by the disposes of image enhancement, binarization, edge detection and removal process noise for the water gauge image. Then it realizes character segmentation based on the characteristics of the characters on the water gauge, achieves character recognition using the template matching method, and uses the longest and other arithmetic method to optimize recognition results. Finally, it calculates water gauge readings based on the recognition results at the different situations. Experimental results show that the correctness and accuracy of the method for identifying information is higher. It may meet the actual application requirements.