Design of on-line nondestructive rapid detection system for tilapia descaling rate based on machine vision

Abstract

Objective: Improve the quality and efficiency of tilapia processing and increase the added value of products. Methods: For the fluorescence response difference of fish scales and fish skin under ultraviolet excitation, a rapid detection system for tilapia descaling rate based on machine vision technology was constructed. After image pre-processing, background segmentation, ROI region selection, and rapid identification of discaled areas, the scaled and descaled area images were obtained. Then the scaling rate was calculated and the detection accuracy of the system was tested. Results: The accuracy of the system scaling rate detection ranged from 90.18% to 94.14%. Conclusion: The automatic transmission of the tilapia sample, the automatic triggering and acquisition of the detection image, the rapid image processing can be realized by the system, which can effectively improve the detection efficiency of tilapia descaling rate and meet the requirements of the production line.

Keywords

• tilapia
• scaling rate
• ultraviolet excitation
• machine vision