AI-based Automated Inspection for Defect Detection

The capability of the machine vision (MV) technology about automatic inspection offers great benefits to the food industry. From simple processes, such as reading bar codes, to the complex process control, tremendous potential for inspection, and imaging that MV endows has brought significant improvisations. Leveraging this technology helps food industrialists to ensure better quality food production in a secure environment.

Our client, a food industry leader, was looking to detect defectively packaged samples on the conveyor belt. PathPartner designed an end-to-end deep learning-based solution to classify the packages as defective and non-defective. This case study talks about:

• Hardware setup and data acquisition
• State-of-the-art hyperparameter optimization (HPO) algorithms to tune models and network customization
• Data pre-processing and visualization for machine learning models
• Data visualization across the pipeline
• Failure case analysis

Download the case study to learn more.