CAJ | 학술논문

　　针对枪弹弹底图像特点，提出一种新的弹底底火表面缺陷图像分割方法。对去背景弹底灰度图像运用Log算子进行边缘检测，分析了传统边界跟踪算法在提取弹底圆轮廓应用中的不足，改进边界跟踪算法提取弹底圆轮廓，利用最小二乘法圆拟合算法获得弹底圆圆心和半径。通过弹底圆半径与底火圆半径对应关系为弹底灰度图像赋值，提取底火圆图像。利用统计阈值分割底火缺陷，利用数学形态学优化分割结果。实验结果表明，该方法能够有效提取弹底底火表面缺陷。
　　침대창탄탄저도상특점，제출일충신적탄저저화표면결함도상분할방법。대거배경탄저회도도상운용Log산자진행변연검측，분석료전통변계근종산법재제취탄저원륜곽응용중적불족，개진변계근종산법제취탄저원륜곽，이용최소이승법원의합산법획득탄저원원심화반경。통과탄저원반경여저화원반경대응관계위탄저회도도상부치，제취저화원도상。이용통계역치분할저화결함，이용수학형태학우화분할결과。실험결과표명，해방법능구유효제취탄저저화표면결함。
According to the properties of cartridge’s bottom, a new method of image segmentation for bullet’s primer surface defect is proposed. The background of cartridge’s bottom image is decreased by the method of background subtraction, it applies Log operator to detect the circle edge of cartridge’s bottom, the disappointment of circle contour extraction by traditional bound-ary tracing is analyzed, the method of boundary tracing is improved to detect the cartridge’s bottom circle, the center and radius of circle are acquired accurately by the least square method. According to the radius relationship between the cartridge’s bottom and primer circle, it evaluates the cartridge’s bottom gray image to extract the image of primer. The primer surface defect is seg-mented by threshold, and the result of segmentation is optimized by mathematical morphology. The result shows this method can extract the bullet’s primer surface defect effectively.