【摘要】 混凝土表面裂缝宽度不仅影响结构物的观瞻,而且影响结构的耐久性。这种裂缝检测问题在国防、航空航天、机械制造等许多领域中都会碰到。因此针对采集到的裂缝图像特征,提出类判别法与数学形态学结合的方法对裂缝进行定量分析。也就是说,根据裂缝图像,利用类判别法找到图像分割的最佳阈值,进而提取图像的裂缝信息,然后通过编制的图像处理程序分析出研究区域的裂缝宽度及其方差。同时,还通过实验对该方法的测量精度进行了验证,结果表明,此方法精度在15%以内,可满足工程需要,因此相关的计算结果可为实际工程的安全可靠性评估提供较准确的基础数据。更多还原

【Abstract】 The surface cracking width obviously affects the appearance of the reinforced concrete structure and the durability of construction.This problem of detecting crack is commonly encountered in many fields,such as:national defence,aviation and space,mechanical fabrication,etc.Aimed at the features of crack image the combined method of category discriminance and mathematical morphology,which is a quantification analysis method,is proposed.According to the crack image,the best threshold is obtained through the analysis method of category discriminance.Based on the threshold,the crack feature is quickly picked up.Furthermore,by taking advantage of the image processing technology,the cracking parameters in this region are also calculated,including the width of every point and the dispersed degree.At the same time,the measurement precision of the method proposed is tested and verified,the results show that the precision is in the limit of 15% and meets the requirement of engineering.The relative results provide the basic data for safety evaluation of civil projects.更多还原