CAJ | 학술논문

针对水下大坝裂缝图像的低信噪比、低对比度以及大坝裂缝形状的不规则性和方向的不确定性等特点导致的检测精确度低、通用性弱的问题，提出了一种基于 Gabor 算子的人工蜂群算法大坝裂缝检测方法。该方法简化了蜜蜂的搜索过程，先利用 Gabor 算子搜索得到一组局部最优花蜜源，根据花蜜源的“收益度”以及“朝向性”特点，从最优花蜜源的角度邻域搜索收益率较高的花蜜源，并对所有的食物源按照条件进行筛选和连接。由于减小了搜索范围，所以可以节省搜索时间，而且针对不同类型的大坝裂缝可以精确连贯地检测出大坝裂缝信息，并且可以有效地抑制噪声。
침대수하대패렬봉도상적저신조비、저대비도이급대패렬봉형상적불규칙성화방향적불학정성등특점도치적검측정학도저、통용성약적문제，제출료일충기우 Gabor 산자적인공봉군산법대패렬봉검측방법。해방법간화료밀봉적수색과정，선이용 Gabor 산자수색득도일조국부최우화밀원，근거화밀원적“수익도”이급“조향성”특점，종최우화밀원적각도린역수색수익솔교고적화밀원，병대소유적식물원안조조건진행사선화련접。유우감소료수색범위，소이가이절성수색시간，이차침대불동류형적대패렬봉가이정학련관지검측출대패렬봉신식，병차가이유효지억제조성。
In this paper,we propose a kind of dam crack detection of artificial colony algorithm based on Gabor operator to solve problems such as low detection precision and application,caused by the low SNR and contrast of the dam crack image,and irregularity and uncertainty of the dam crack shape. This method simplifies the search process of bees.First,the Gabor operator is used to search the edge of image to get a set of local points and its direction.And then,according to the characteristics of its revenue degree and orientation,the food source is searched in the optimal food source neighborhood.At last,the image is searched from the local optimal points and all dam cracks information is found.We can save search time due to reducing the search scope and detect the dam crack information consistently according to the different types of dam crack.Moreover,it can suppress noise effectively.