Abstract :

The Quality Control Of Biscuits In Large-scale Food Manufacturing Is A Critical Factor In Ensuring Customer Satisfaction And Brand Reputation. Traditional Manual Inspection Methods Are Time-consuming, Prone To Human Error, And Lack Consistency, Particularly When Detecting Subtle Defects. To Address These Limitations, This Research Proposes A AI-Based Biscuit Defect Detection System For Enhancing Food Industry Quality Control Is Designed To Automatically Identify And Classify Biscuit Defects Using Machine Learning And Deep Learning Techniques. In Traditional Food Manufacturing Industries, Biscuit Quality Inspection Is Typically Performed Manually, Which Can Be Time-consuming, Inconsistent, And Prone To Human Error. To Address This Challenge, The Proposed System Utilizes Image Processing And Artificial Intelligence To Analyze Biscuit Images And Detect Defect Categories Such As Color Defect, No Defect, Object Defect, And Shape Defect. The Dataset Is First Preprocessed Through Image Resizing, Normalization, And Conversion Into Numerical Arrays To Prepare It For Model Training. Multiple Classification Models Including Extra Trees Classifier (ETC), Linear Discriminant Analysis (LDA), And Light Gradient Boosting Machine Classifier (LGBM) Are Implemented And Compared With A Proposed ResNet-based Convolutional Neural Network (CNN) Model That Automatically Extracts Complex Visual Features From Biscuit Images. The Models Are Evaluated Using Performance Metrics Such As Accuracy, Precision, Recall, And F1-score, Along With Visualization Techniques Including Confusion Matrices And ROC-AUC Curves To Measure Classification Effectiveness. Experimental Results Demonstrate That The ResNet CNN Model Achieves Superior Performance With An Accuracy Of Approximately 98.98%, Significantly Outperforming Traditional Machine Learning Models. The System Is Implemented With A User-friendly Graphical Interface That Supports Dataset Management, Model Training, Prediction, And Result Visualization, Along With Additional Capabilities Such As Batch Prediction And Telegram Bot Integration For Remote Biscuit Quality Analysis. The Proposed System Provides An Efficient, Automated, And Reliable Solution For Biscuit Defect Detection, Helping Food Industries Improve Product Quality Control, Reduce Manual Inspection Efforts, And Ensure Consistent Manufacturing Standards Through Intelligent Image-based Analysis.

Published:

24-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

3035-3044

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite