Development of Yarn Quality Evaluation and Prediction Model Using Feature Extraction and Neural Network-Based Evaluation in Image Processing

Abstract

Yarn quality prediction is an important aspect in the textile industry that plays a significant role in optimizing the performance of spinning mills. In which proposed method to use to reducing the computational cost and improving the production efficiency. The study employs a neural network-based BPNN algorithm for classifying and extracting prominent features for predicting yarn quality. The performance analysis of the proposed approach has been conducted by evaluating the performance parameters of the yarn quality. The prediction models such as Yarn length, Hairiness length, Fuzzy Yarn length, and Fuzzy Hairiness length used for evaluation. A comparative analysis stayed conducted to validate the performance of the approach. It is done by comparing the results of the proposed frameworks with another efficient neural network MR-MRF method. Results show the effectiveness of the recommended methodology by achieving superior precision and accuracy.