Assessing The Impact Of CNN Auto Encoder-Based Image Denoising On Image Classification Tasks

Images captured from the real world are often affected by different types of noise, which can significantly impact the performance of Computer Vision systems and the quality of visual data. This study presents a novel approach for defect detection in casting product noisy images, specifically focusing on submersible pump impellers. The methodology involves utilizing deep learning models such as VGG۱۶, InceptionV۳, and other models in both the spatial and frequency domains to identify noise types and defect status. The research process begins with preprocessing images, followed by applying denoising techniques tailored to specific noise categories. The goal is to enhance the accuracy and robustness of defect detection by integrating noise detection and denoising into the classification pipeline. The study achieved remarkable results using VGG۱۶ for noise type classification in the frequency domain, achieving an accuracy of over ۹۹%. Removal of salt and pepper noise resulted in an average SSIM of ۸۷.۹, while Gaussian noise removal had an average SSIM of ۶۴.۰, and periodic noise removal yielded an average SSIM of ۸۱.۶. This comprehensive approach showcases the effectiveness of the deep AutoEncoder model and median filter, for denoising strategies in real-world industrial applications. Finally, our study reports significant improvements in binary classification accuracy for defect detection compared to previous methods. For the VGG۱۶ classifier, accuracy increased from ۹۴.۶% to ۹۷.۰%, demonstrating the effectiveness of the proposed noise detection and denoising approach. Similarly, for the InceptionV۳ classifier, accuracy improved from ۸۴.۷% to ۹۰.۰%, further validating the benefits of integrating noise analysis into the classification pipeline.

کلیدواژه ها:

Image denoising ، CNN AutoEncoder ، Image classification ، Transfer learning ، industrial defect inspection