DEGP-YOLOV10: A YOLO Model for Table Row and Column Detection

Since the invention of double-entry bookkeeping, reading complex financial statements and cash flow accounts has been considered a complex task that requires a large number of professionals. With the development of artificial intelligence, it seems possible to extract key information from tables through auxiliary tools to improve the efficiency of completing such tasks and reduce the corresponding human resource requirements. This paper proposes a solution based on the CSPNet yolo v10 model to improve the efficiency of identifying and extracting key items in large tables. Compared with the basic CSPNet yolo v10 model, this new model has stronger detail capture and global perception capabilities, stronger detail capture and semantic fusion capabilities, and better accuracy and convergence speed.