基于百度飞桨EasyDL平台的芯片质检系统设计与开发
打开文本图片集
中图分类号：TP311 文献标识码：A 文章编号：2096-4706（2025）12-0184-10
Design and Development of Chip Quality Inspection System Based on PaddlePaddle EasyDLPlatform
SUN Cuigai， SHENG Xuefeng，YU Dawei （Suzhou CollegeofInformationTechnology，Suzhou 21520o，China）
Abstract：Aimingat the problems oflowchip inspection eficiency，high labor cost，and poor product consistency in current industrialproduction，this paper designsand develops anautomatedchipquality inspection system basedonartifcial intellgencetechnology.ThesystemtakesthePaddePaddleEasyDLplatformas thecore，colectschipimages throughindustrial cameras，anduses the Deep Learning algorithmtorealize the inteligentrecognition of defectssuchaschippinmising and surfacecratches.Theexperimentalresultsshowthatthesystemcanefectivelyreplace thetraditionalmanual inspectionmethod， significantlyimprovetheispectionaccracyandeiencyedueproductionosts，andprovidestrongsupportforancing thecorecompetitivenssof theindustryandpromotingkeytechnologicalinnovations.Thesuccessfulapplicationofthissystem notonlypromotes theimprovementofindustrial independentinnovationabilityutalsoprovidesanetechnicalpathforthe realization of high-quality development of manufacturing industry.
Keywords： PaddlePaddle; EasyDL; chip inspection; quality inspection; Deep Learning; intellgent manufacturing
0 引言
在现代电子产业中，芯片作为核心组件，其质量直接关系到电子产品的性能、稳定性和可靠性。（剩余10289字）