Date of Publication :30th August 2023

Abstract: In many manufacturing processes, quality control is the main component specially the casting and welding. The manual procedures of the quality control is time consuming, so to meet the high-quality product images demand, the utilization of the visual inspection systems is becoming more attractive. Different Artificial intelligence techniques showed outstanding performance in classification, detection and localization tasks. Network is training simultaneously for the detection of defects and segmentation resulting high accuracy. In manufacturing, product image defect detection is important and in this paper, different existing Machine Learning (ML) and Deep learning (DL) methods for detection of defects are surveyed. The product image defects are classified and different available techniques are discussed in this paper for the defect detection. Feature analysis and segmentation techniques are also surveyed. This paper summarize various Machine learning and deep learning techniques for defect detection. The equipment’s functions and characteristics used for the detection of defects are also investigated.

Reference :