نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Accurate identification of the car body platform before entering the paint spray cabin is vital to prevent costly errors, optimize material consumption, and increase reliability in the automotive production line. Conventional systems based on optical sensors lack the required accuracy in the complex industrial environment due to their inability to distinguish between similar bodies and perform real-time processing. This research focuses on the design and implementation of an intelligent system based on image processing and deep learning that is capable of automatically and accurately identifying three different car body platforms (X200, SP100, and optimized X200). In the proposed architecture, the ResNet backbone along with transfer learning and advanced data augmentation techniques was employed to ensure adaptability and robustness against variable industrial conditions (lighting variations and background complexity). The model was trained and extensively evaluated on a comprehensive and real dataset collected from the production line. Experimental results showed that the ResNet-50 model achieved 100% recognition accuracy and an average confidence level above 98% in a real and challenging operational environment. This system possesses high stability and reliability and has full potential to replace traditional systems in order to reduce errors, save material consumption, and minimize dependence on human intervention in the car painting line.
کلیدواژهها English
