Comparative Study of Novel Convolutional Neural Network and Artificial Neural Network to Enhance the Accuracy for Flaw Detection in Animal Hide

This research explores how well Artificial Neural Networks (ANN) and Convolutional Neural Networks (NCNN) detect animal hide imperfections to improve quality of weather. Materials and Methods: The study extracted 3600 leather images from Kaggle and split them into two separate groups for this investigation. Group 1 used NCNN, and Group 2 used ANN. Statistical testing generated a sample size of 20 specimens at 95% confidence with 80% power in this experiment. Python became the chosen programming platform for data analysis and preprocessing. NCNN shows better results than ANN because it delivers 96.80% accuracy at 0.001 significance level (p<0.05). Conclusion: The analysis proves the Novel Convolutional Neural Network successfully finds defects in animal hides, but ANN cannot match its results.