CAJ | 학술논문

Canny算子因其信噪比高、定位准确以及单边响应的优势，常常用于图像的边缘检测，而双阈值检测中阈值的选取会影响图像边缘提取的效果，为了获得边缘检测的最佳阈值，引入了一种改进的遗传算法。该方法从遗传算子和操作策略两个方面对基本遗传算法作出改进，重点设计了种群进化的适应度函数，使用该方法确定图像边缘连接的最佳阈值，以获得图像的边缘检测效果图。仿真结果表明：采用改进的遗传算法确定的阈值所得到的图像，边缘细节丰富，定位准确。
Canny산자인기신조비고、정위준학이급단변향응적우세，상상용우도상적변연검측，이쌍역치검측중역치적선취회영향도상변연제취적효과，위료획득변연검측적최가역치，인입료일충개진적유전산법。해방법종유전산자화조작책략량개방면대기본유전산법작출개진，중점설계료충군진화적괄응도함수，사용해방법학정도상변연련접적최가역치，이획득도상적변연검측효과도。방진결과표명：채용개진적유전산법학정적역치소득도적도상，변연세절봉부，정위준학。
Due to the advantages of high signal-to-noise ratio,accurate positioning and unilateral response,the canny operator is used in image edge detection,and the threshold of double thresh-old detection can affect the image edge extraction effect.In order to acquire the best threshold,an improved genetic algorithm is adopted.The basic genetic algorithm is improved from two aspects of genetic operators and operating strategy,focusing on the fitness function design of the popula-tion evolution,which is used to determine the best threshold of image edge connection,so that the image edge detection effect is obtained.Results show that the threshold determined by the im-proved genetic algorithm is used to image edge detection,which can obtain more details and accu-rate positioning.