基于计算机视觉的汽车零部件品质检测与评估

打开文本图片集

Quality Inspection and Evaluationof Automotive Parts Based on Computer Vision

Chen Xiaodong

（HenanIndustry and Trade Vocational College，Zhengzhou ，China）

【Abstract】Aiming at the problem of parts quality inspection in automobile production，this article proposes a solution based on computer vision technology.Firstly，construct the basic architecture of the detection system including imageacquisitionhardware，imageprocessing algorithmsand featureparametercalculation.Among them，thehardware adoptscomponents such as ring-shaped LED light sources and CMOS industrial cameras，and thealgorithm realizes defect detection through preprocesing，featureextraction，etc.Secondly，a multi-dimensional evaluation index system forgeometricaccuracy，surfacequalityand comprehensive performanceis established.Thegeometricaccuracy evaluation adopts amulti-scale fusionarchitecture and improvedalgorithms.The surfacequality evaluation combines multi-mode imagesand frequency-domain analysis，and thesystemacuracyisverified through processes such asreference calibrationanderortraceability.Thestabilityisevaluatedusingamulti-physicalfieldcoupling testchamber.Finall，a systematicverificationcapabilityfromstaticcalibrationtofullworkingconditioncoverageisstudiedandformed， providingan operational test optimizationsolution for automotive manufacturing enterprisesand promoting the transformation of the detection system towards "autonomous evolution".

【Key words】 visual technology；automobile parts；quality inspection；stability assessment

0 引言

汽车生产中，零件品质是决定汽车产品性能的关键指标。（剩余4011字）