DEFECT INSPECTION SYSTEM AND DEFECT INSPECTION METHOD

• Main Author: WATANABE KIMIHIKO
• Format: Patent
• Language: English; Japanese
• Published: 15.12.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIRCHEMICAL OR PHYSICAL PROPERTIES; MEASURING; PHYSICS; TESTING

To provide a defect inspection system and a defect inspection method, reducing excessive detection of good articles and preventing an outflow of defective articles.SOLUTION: A defect inspection system 100 determines a quality of an inspection object 1 according to inspection target image data 75. The defect inspection system 100 includes: an inspection processing part 146 using a learned model configured by multiple pieces of good article image data 40 and multiple pieces of defective article image data 30 to output a score value of an inspection target image data 75 to execute quality determination; a first determination part 147 configured to determine whether or not the score value is within a range 86 of a third threshold value 85 and a first threshold value 83; a determination sensitivity introduction part 148 configured to correct the score value by multiplying with determination sensitivity if the score value is within the range 86; and a second determination part 149 configured to determine whether the corrected score value is equal to or more than the first threshold value 83. The inspection processing part 146 determines as a defective article when the corrected score value is equal to or more than the first threshold value 83 and determines as a defect being uncertain if the corrected score value is less than the first threshold value 83. A defect inspection method is also provided.SELECTED DRAWING: Figure 10 【課題】良品の過検出を低減し不良品の流出を抑制する欠陥検査システム及び欠陥検査方法を提供する。【解決手段】本発明は、検査対象画像データ７５に基づき検査対象物１の良否判定する欠陥検査システム１００であって、複数の良品画像データ４０及び複数の不良品画像データ３０により構築された学習済みモデルを用いて、検査対象画像データ７５のスコア値を出力し良否判定する検査処理部１４６と、スコア値が第３閾値８５と第１閾値８３との範囲８６内か判定する第１判定部１４７と、スコア値が範囲８６内の場合に判定感度を乗じてスコア値を補正する判定感度導入部１４８と、補正スコア値が第１閾値８３以上か判定する第２判定部１４９と、を備え、検査処理部１４６は、補正スコア値が第１閾値８３以上の場合に不良品判定し、補正スコア値が第１閾値８３未満の場合に判定不定とする欠陥検査システム１００及び欠陥検査方法に関する。【選択図】図１０