System and method for crispness determination and grading prediction of crisp tilapia based on texture analyzer and neural network

Abstract

The invention discloses a system and method for crispness determination, grading prediction of crisp tilapia based on a texture analyzer, a neural network. The system includes: a data acquisition module, a model training module, a grading prediction module; the data acquisition module is used for sampling crisp tilapia to obtain test samples, performing crispness detection, grade scoring on the test samples to obtain test sample evaluation results; the model training module is used for constructing a grade division model based on a BP neural network, preprocessing the test sample evaluation results, inputting the preprocessed results into the grade division model for training to obtain a trained grade division model; the grading prediction module is used for randomly selecting a preset number of crisp tilapia samples from the test samples, and inputting the crisp tilapia samples into the trained grade division model to obtain tilapia crispness determination and grading prediction results.