Silkscreen Text Identification of Circuit Card Assemblies by Optical Character Recognition for AutoBOM Extraction

AsterTech proposes an automated optical character recognition (OCR) technique to extract Bill of Materials (BOM) information of circuit card assemblies (CCA) from their optical images. Our method will recognize texts in the optical image by appropriate machine learning models based on their appearance, characters and trained data-set complemented with prior domain knowledge. These recognized markings will allow easy identification of CCA components such as resistors, capacitors, ICs, inductor, no component, etc. for comprehensive retrieval of BOM from a complete non-destructive analysis process.