计算机工程与设计
計算機工程與設計
계산궤공정여설계
COMPUTER ENGINEERING AND DESIGN
2014年
7期
2434-2437,2475
,共5页
颜色缺陷%图像处理%高斯金字塔分解%互相关匹配%CIELAB空间%CIEDE2000色差
顏色缺陷%圖像處理%高斯金字塔分解%互相關匹配%CIELAB空間%CIEDE2000色差
안색결함%도상처리%고사금자탑분해%호상관필배%CIELAB공간%CIEDE2000색차
color defects%image processing%Gaussian pyramid decomposition%cross-correlation matching%CIELAB color space%CIEDE2000 color difference
为了实现遥控器颜色缺陷的自动检测,进行了遥控器缺陷检测以及印刷缺陷检测的调查,提出了一种基于图像处理的自动检测算法。针对传统的互相关匹配速度慢的缺点,采用高斯金字塔分解与互相关匹配结合的方法对样品图像和模板图像进行匹配,保证了匹配速度和精度;将两幅图像转换到与人的目视感觉最符合的均匀色彩空间CIE1976 L*a*b*(简称CIELAB),利用色差公式CIEDE2000来计算二者的色差值;依据色差值来区分正常区域和缺陷区域,得出最终检测结果。实验结果表明,该算法能很好地检测出遥控器的颜色缺陷。
為瞭實現遙控器顏色缺陷的自動檢測,進行瞭遙控器缺陷檢測以及印刷缺陷檢測的調查,提齣瞭一種基于圖像處理的自動檢測算法。針對傳統的互相關匹配速度慢的缺點,採用高斯金字塔分解與互相關匹配結閤的方法對樣品圖像和模闆圖像進行匹配,保證瞭匹配速度和精度;將兩幅圖像轉換到與人的目視感覺最符閤的均勻色綵空間CIE1976 L*a*b*(簡稱CIELAB),利用色差公式CIEDE2000來計算二者的色差值;依據色差值來區分正常區域和缺陷區域,得齣最終檢測結果。實驗結果錶明,該算法能很好地檢測齣遙控器的顏色缺陷。
위료실현요공기안색결함적자동검측,진행료요공기결함검측이급인쇄결함검측적조사,제출료일충기우도상처리적자동검측산법。침대전통적호상관필배속도만적결점,채용고사금자탑분해여호상관필배결합적방법대양품도상화모판도상진행필배,보증료필배속도화정도;장량폭도상전환도여인적목시감각최부합적균균색채공간CIE1976 L*a*b*(간칭CIELAB),이용색차공식CIEDE2000래계산이자적색차치;의거색차치래구분정상구역화결함구역,득출최종검측결과。실험결과표명,해산법능흔호지검측출요공기적안색결함。
To realize the automatic detection of color defects on remote keypad,the remote control defects detection and printing defects detection methods were investigated,then an automated defects detection algorithm based on image processing was pro-posed.To overcome the shortcoming of slow speed of the traditional cross-correlation matching,and algorithm combined the Gaussian pyramid decomposition and the cross-correlation matching,and performed on the model image and the sample image, giving a good tradeoff between the efficiency and execution time.Then the two images were transformed into the CIE1976L *a*b color space,which was uniform and conform to human visual sense.Then the color difference of the two images was com-puted by CIEDE2000.According to the color difference value,the defective regions could be distinguished from the normal re-gions and the final results were achieved.The results of simulation experiments showed that the proposed algorithm could detect the color defects on the remote keypad well.