CAJ | 학술논문

为提高视觉导航AGV在运行过程中对路径标识符识别的可靠性和实时性，首先设计了实际应用中常见的几类标识符，然后针对这些标识符提出了一种基于形状特征的识别算法，最后通过实验验证了该算法的可行性，并对实验结果进行了分析，分析表明当标识符存在污损、旋转、比例缩放以及光照条件变化等情况时，该算法仍然能够获得快速高精度的识别效果。
위제고시각도항AGV재운행과정중대로경표식부식별적가고성화실시성，수선설계료실제응용중상견적궤류표식부，연후침대저사표식부제출료일충기우형상특정적식별산법，최후통과실험험증료해산법적가행성，병대실험결과진행료분석，분석표명당표식부존재오손、선전、비례축방이급광조조건변화등정황시，해산법잉연능구획득쾌속고정도적식별효과。
In order to obtain better robustness and real -time during the recognition of path identifiers at the vi-sion-based automatic guided vehicle (AGV) runs, it designs several kinds of common identifiers in practical ap-plication .And then aiming at these identifiers , it proposes a method of identifiers recognition based on shape fea-tures, shows the detail about extractting the image edges with Canny operator , construction the histogram of ori-ented gradient .It uses Euclidean distance as similarity measurement to match the model image and test image . Experimental results demonstrate that the method has fast speed and high accuracy when the identifiers are dirty , rotated, scale-changed or the illumination changes .