Abstract: The design and implementation of an intelligent system that uses machine learning and image processing techniques to identify wear and tear in automotive components is presented in this paper. From image acquisition and validation to feature extraction, classification, and result visualization, the suggested solution makes use of an intuitive MATLAB graphical user interface (GUI). Following validation and pre-processing, Gabor filters are used to extract pertinent texture features that indicate wear patterns from images of automotive parts. A Probabilistic Neural Network (PNN) is then used to classify these features, dividing the component's condition into three different categories according to severity and estimated age. The system's outputs, which include component type, quality, and age, are clear and actionable, allowing for prompt maintain decision. The system's ability to automate the diagnostic process and achieve high accuracy in classifying wear levels is demonstrated by extensive testing. Reliability is increased by robust error handling and data validation, while accessibility for users with little technical expertise is guaranteed by the modular GUI design. This study lays the groundwork for future improvements utilizing deep learning and bigger datasets while demonstrating the potential of fusing machine learning and image analysis for useful, real-world applications in automotive maintenance.

Keywords: Image Processing, Machine Learning, Gabor Filters, Probabilistic Neural Network (PNN), Feature Extraction, MATLAB GUI.

| DOI: 10.17148/IARJSET.2025.125170