CAJ | 학술논문

针对传统边缘检测算法自适应能力差、固定阈值和背景噪声抑制差的问题，提出了一种基于萤火虫算法的图像边缘测算法。该算法首先通过灰度图像矩阵的一阶导数得到灰度图像的梯度值矩阵，然后使用萤火虫算法，即根据萤火虫的发光和吸引特性，根据其搜索区域范围寻找其它萤火虫，并向邻域内位置内亮度高的萤火虫位置移动，从而实现搜索图像的梯度最大值，检测出图像的边缘。仿真实验结果表明，该算法能快速、准确地检测出图像的边缘，因此将其应用于图像的边缘检测是可行、有效的。
침대전통변연검측산법자괄응능력차、고정역치화배경조성억제차적문제，제출료일충기우형화충산법적도상변연측산법。해산법수선통과회도도상구진적일계도수득도회도도상적제도치구진，연후사용형화충산법，즉근거형화충적발광화흡인특성，근거기수색구역범위심조기타형화충，병향린역내위치내량도고적형화충위치이동，종이실현수색도상적제도최대치，검측출도상적변연。방진실험결과표명，해산법능쾌속、준학지검측출도상적변연，인차장기응용우도상적변연검측시가행、유효적。
Aiming at the deficiencies of low self-adaptive abilities, fixed threshold and weak background noise suppression, an image edge detection algorithm based on GSO ( Glowworm Swarm) is proposed for the first time. This algorithm first acquires the gradient matrix of grayscale by first-order derivative of grayscale matrix,then, glowworm swarm algorithm is used to find other glowworm according to flashing and attracting characteristics of the glowworm and the region of search,thus moving towards the luminant glow-worm so as to achieve the maximum of the searched image gradient and detect the image edge. Simulation experiment indicates that this algorithm could detect the image edge rapidly and accurately,and thus is fea-sible and effective for its application in image edge detection.