CAJ | 학술논문

结合蚁群反馈优化算法、多属性约束理论的优点，提出了一种基于约束蚁群反馈算法的钢坯图像分割算法。算法采用钢坯图像边界粗糙熵作为分割的反馈标准，充分利用蚁群算法的反馈调节能力，对属性像素进行约束，保证在约束条件下的最优化边沿分割。经过实验测试，结果表明，由于使用约束蚁群反馈算法避免了钢坯像素边沿混合弊端，采用模糊边界计算原则，完成边沿像素分割，所以不仅得到了令人满意的钢坯分割效果，而且大大的提高了计算速度，是一种实用有效的钢坯图像分割方法。
결합의군반궤우화산법、다속성약속이론적우점，제출료일충기우약속의군반궤산법적강배도상분할산법。산법채용강배도상변계조조적작위분할적반궤표준，충분이용의군산법적반궤조절능력，대속성상소진행약속，보증재약속조건하적최우화변연분할。경과실험측시，결과표명，유우사용약속의군반궤산법피면료강배상소변연혼합폐단，채용모호변계계산원칙，완성변연상소분할，소이불부득도료령인만의적강배분할효과，이차대대적제고료계산속도，시일충실용유효적강배도상분할방법。
This article feedback combined with ant colony optimization algorithm, the advantages of multiple attribute con-straint theory, proposed an ant colony based on constraint feedback billet image segmentation algorithm of the algorithm. Algorithm USES the billet image boundary rough entropy as the division standard of feedback, and it makes full use of the feedback adjustment ability of ant colony algorithm, constraints on properties of pixels. Results show that the feed-back due to the use of ant colony algorithm to avoid the drawbacks of billet mixed pixel edge, fuzzy boundary calculation principles, complete division of edge pixels, so not only got a satisfactory segmentation results of billet, and greatly im-proves the computing speed, is a practical and efficient billet image segmentation method.