AN INTEGRATED FMEA–QFD AND IMAGE PROCESSING APPROACH FOR OVARIAN MATURITY CLASSIFICATION IN CRAB PROCESSING SYSTEMS

Abstract

Mud crabs are high-value products in Thailand’s seafood industry, though ovarian maturity significantly influences market price. In current production practice, ovarian maturity assessment relies on invasive and expert-dependent methods, resulting in variability, inaccuracy, and operational inefficiency. These constraints limit process standardization and industrial scalability. Therefore, this study aimed to develop a non-invasive crab ovarian maturity classification process to enhance assessment reliability, improve process consistency, and replace expert-dependent evaluation without causing harm to the crabs. The first step involved creating image acquisition equipment based on Failure Modes and Effects Analysis (FMEA) and Quality Function Deployment (QFD) principles. Subsequently, ovarian maturity stages were automatically classified using an image processing technique. The results of the integrated FMEA-QFD and image processing process demonstrated that ovarian maturity can be classified into four stages: developing, early maturing, late maturing, and fully maturing. The main categorization parameter was the ratio of the ovarian area to the carapace area, which increased with crab maturation. The ratio of full maturation was at least 60%, equivalent to premium market value. The approach attained an accuracy of 90% in comparison to expert evaluation. This innovative, non-invasive process for crab ovarian maturity classification can improve consistency and quality control, eliminate specialized workers, and ensure the crabs remain unharmed.