激光标记自动跟踪视频引伸计的标记识别算法研究

摘要

Aiming at the problem that conventional video extensometers are usually susceptible to light changes ,this paper comes up with an image preprocessing algorithm based on combination of image bright-ness adjustment and defogging enhancement .Through such algorithm ,the clear image with laser mark and specimen mark can be obtained .The laser mark and sample mark can be extracted through Cr chrominance threshold segmentation algorithm and brightness gradient threshold segmentation algorithm based on area of interest .The experiments are carried out respectively under light changes and chuck interference of test machine to verify the stability and accuracy of the algorithm .%针对视频引伸计易受光照变化干扰的问题，提出一种基于图像亮度调整和去雾增强相结合的图像预处理算法，获得了激光标记和试样标记的清晰图像；分别利用 Cr色度阈值分割算法和基于感兴趣区域的亮度梯度阈值分割算法提取激光标记和试样标记。在光照变化和有无试验机夹头干扰的情况下分别进行了实验，验证了算法的稳定性和准确性。