农机化研究
農機化研究
농궤화연구
JOURNAL OF AGRICULTURAL MECHANIZATION RESEARCH
2015年
7期
70-73
,共4页
高斯拉普拉斯算子%马铃薯%干腐%结痂%缺陷检测%B通道灰度化%迭代阈值分割法
高斯拉普拉斯算子%馬鈴藷%榦腐%結痂%缺陷檢測%B通道灰度化%迭代閾值分割法
고사랍보랍사산자%마령서%간부%결가%결함검측%B통도회도화%질대역치분할법
gaussian Laplacian operator%potato%dry rot%scab%defect detection%B-channel grayscale%iterative thresh-old segmentation method
提出一种用高斯拉普拉斯算子的方法对马铃薯斑点缺陷进行检测，也称为“LoG斑点”检测，该算法结合B通道灰度化及迭代阈值分割法可以有效地识别马铃薯表皮的干腐、结痂缺陷。实验采集了258个马铃薯样本的516个正反面图像，正确识别率94％以上。结果表明：该方法对检测马铃薯的干腐及结痂缺陷有效。
提齣一種用高斯拉普拉斯算子的方法對馬鈴藷斑點缺陷進行檢測，也稱為“LoG斑點”檢測，該算法結閤B通道灰度化及迭代閾值分割法可以有效地識彆馬鈴藷錶皮的榦腐、結痂缺陷。實驗採集瞭258箇馬鈴藷樣本的516箇正反麵圖像，正確識彆率94％以上。結果錶明：該方法對檢測馬鈴藷的榦腐及結痂缺陷有效。
제출일충용고사랍보랍사산자적방법대마령서반점결함진행검측，야칭위“LoG반점”검측，해산법결합B통도회도화급질대역치분할법가이유효지식별마령서표피적간부、결가결함。실험채집료258개마령서양본적516개정반면도상，정학식별솔94％이상。결과표명：해방법대검측마령서적간부급결가결함유효。
This paper presents a potato spot defects detection can effectively identify the potato skin using Gaussian Laplacian operator, also known as“LoG spot” test, the algorithm is combined with the B-channel grayscale and iterative threshold segmentation method can detect the dry rot, scab defect.Experiment collected 258 potato samples and gathered 516 front and back images of the potato samples,correct identification rate of 94%or more, the results indicate that the method for detection of potato dry rot, scab defects is effective.
