In this research, the performance of two advanced object detection algorithms, YOLOv7 and YOLOv10, were evaluated for accurate defect detection in ceramic tiles. Using a comprehensive dataset of diverse defective tile images, these algorithms were trained and tested on their ability to precisely identify various defects such as edge chipping, holes, and surface scratches. The results demonstrated that YOLOv10 significantly outperformed YOLOv7, exhibiting a higher capability for detecting a wider range of defects.Our findings highlight the substantial potential of deep learning algorithms like YOLOv10 for industrial inspection applications. Compared to traditional inspection methods, which are often time-consuming, costly, and prone to human error, deep learning algorithms can detect defects with remarkable speed and accuracy. This leads to significant reductions in production costs, increased efficiency, and higher product quality. Moreover, the early detection of defects by these algorithms prevents more serious issues and associated costs, ultimately improving the overall production process.However, it's important to note that this research focused specifically on ceramic tile defects. Further investigations are required to generalize the findings to other materials and industries.