Innovative Detection Method of Wood Cracks Using YOLO11 and Morphological Processing

Wood is a historically significant building material that has been used worldwide for thousands of years. As a construction material, wood is primarily subject to brittle failure, and the instability of wooden structures poses a threat to their structural integrity, durability, and safety. Therefore, effective detection of cracks in wooden structures is crucial. This paper focuses on the identification of wood damage cracks and establishes a novel and diverse datasets specifically for this purpose. For the first time, we combine morphological opening operations in image processing with YOLO11 to optimize detection in complex crack scenarios, proposing a framework called TimberCrackYolo for wood crack detection. By selecting the most suitable kernel size for the opening operation, this framework achieves a Precision of 81.6% and a Recall rate of 84.3% in damage crack identification, improving performance by 1–5% over the existing YOLO11 model. This method is more intelligent and efficient compared to current detection techniques, having a profound impact on the safety assessment of wooden structures.