武汉工程大学学报
武漢工程大學學報
무한공정대학학보
Journal of Wuhan Institute of Technology
2015年
11期
43-46
,共4页
王学华%王华龙%马凡杰%李安翼%刘苏
王學華%王華龍%馬凡傑%李安翼%劉囌
왕학화%왕화룡%마범걸%리안익%류소
字符识别%HALCON%图像处理%字符图像分割
字符識彆%HALCON%圖像處理%字符圖像分割
자부식별%HALCON%도상처리%자부도상분할
character recognition%HALCON%image processing%character image segmentation
在大规模工业生产过程中,几乎所有产品表面都会负载产品标示的字符信息。采取字符识别的方式对产品信息进行自动采集及判别在产品的整个生命周期中起着重要作用,因此满足工业现场工况要求的字符识别系统必须具备识别速度快,辨识效率高﹑鲁棒性好的特点。基于机器视觉平台MVTec HAL-CON,结合高级语言编程,设计和开发了铜环产品柱面喷码信息的自动采集及在线识别系统,包含物料输送﹑图像采集﹑信息判断及分拣,解决了工业铜环柱面不规则排列字符﹑扭曲变形字符和模糊字符信息的图像处理和OCR(Optica1 Character Recognition)字符识别问题。运行结果表明不规则字符信息一次识别率达到99%以上,实现了铜环产品的生产与检测过程一体化。
在大規模工業生產過程中,幾乎所有產品錶麵都會負載產品標示的字符信息。採取字符識彆的方式對產品信息進行自動採集及判彆在產品的整箇生命週期中起著重要作用,因此滿足工業現場工況要求的字符識彆繫統必鬚具備識彆速度快,辨識效率高﹑魯棒性好的特點。基于機器視覺平檯MVTec HAL-CON,結閤高級語言編程,設計和開髮瞭銅環產品柱麵噴碼信息的自動採集及在線識彆繫統,包含物料輸送﹑圖像採集﹑信息判斷及分揀,解決瞭工業銅環柱麵不規則排列字符﹑扭麯變形字符和模糊字符信息的圖像處理和OCR(Optica1 Character Recognition)字符識彆問題。運行結果錶明不規則字符信息一次識彆率達到99%以上,實現瞭銅環產品的生產與檢測過程一體化。
재대규모공업생산과정중,궤호소유산품표면도회부재산품표시적자부신식。채취자부식별적방식대산품신식진행자동채집급판별재산품적정개생명주기중기착중요작용,인차만족공업현장공황요구적자부식별계통필수구비식별속도쾌,변식효솔고﹑로봉성호적특점。기우궤기시각평태MVTec HAL-CON,결합고급어언편정,설계화개발료동배산품주면분마신식적자동채집급재선식별계통,포함물료수송﹑도상채집﹑신식판단급분간,해결료공업동배주면불규칙배렬자부﹑뉴곡변형자부화모호자부신식적도상처리화OCR(Optica1 Character Recognition)자부식별문제。운행결과표명불규칙자부신식일차식별솔체도99%이상,실현료동배산품적생산여검측과정일체화。
The character information about product ID wi11 be attached on the surfaces of a1most a11 of the products in the 1arge-sca1e industria1 production process, and the automatic co11ection and identification of product ID in the way of character recognition p1ays an important ro1e in the who1e 1ife cyc1e of the products. To meet the workshop working app1ications, the character recognition system must have the characteristics of rapid recognition speed, high identification efficiency and good robustness. Combined with advanced language programming,an automatic data collection and online recognition system for the information printed on the dylinder surface copper ring product was proposed based on MVTec HALCON machine vision platform. The proposed system includes parts automated de1ivery, image acquisition, information judgment and sorting, in which the prob1ems of image processing and optica1 character recognition(OCR) of characters with irregu1ar arrangement, distortion and b1ur on the cy1inder surface of copper ring was so1ved. The resu1t indicates that the recognition rate of the irregu1ar character information is up to 99%, and the integration of products pro-ducing and detecting process is rea1ized for the copper ring production.