CAJ | 학술논문

图像匹配技术是图像测量系统中的关键技术，出于对目前常用的两种图像匹配模型的研究，将基于图像几何特征的匹配模型和基于频域的匹配模型结合汽车车架上大梁的机器视觉测量的工程实例，通过小旋转角和大旋转角两组数据以及编程实现两者的对比，阐述了两种方法的匹配精度、误差来源，并得出基于几何特征的匹配模型在大梁匹配定位中有更高精度。
도상필배기술시도상측량계통중적관건기술，출우대목전상용적량충도상필배모형적연구，장기우도상궤하특정적필배모형화기우빈역적필배모형결합기차차가상대량적궤기시각측량적공정실례，통과소선전각화대선전각량조수거이급편정실현량자적대비，천술료량충방법적필배정도、오차래원，병득출기우궤하특정적필배모형재대량필배정위중유경고정도。
Image matching is the key technology of image measurement system, based on two kinds of image matching model of commonly used at present, will be based on the matching model of geometric characteristic of image matching based on frequency domain and model, vehicle frame crossbeam of machine vision detection combined with engineering instance, the contrast by small angle and large rotation corner two group and programming, the matching accuracy two methods of error sources, and geometric characteristics of the matching model in beam matching more accurate positioning.