Construction site scaffold spacing detection method based on computer vision

The invention relates to a construction site scaffold spacing detection method based on computer vision. The method comprises the following steps: firstly, extracting skeleton information of a scaffold image by adopting a DeepFlux skeleton extraction algorithm; secondly, after a scaffold skeleton image is obtained, according to the special physical structure of the scaffold, an intersection point detection method is adopted to calculate the intersection point of the scaffold; and finally, according to the obtained intersection point, calculating parameter information of the scaffold image by adopting Euclidean distance, and converting the image data into actual data by adopting a target method so as to judge whether the mounting distance of the scaffold is standard or not. Compared with a traditional outdated method for manually measuring the scaffold, the method has the advantages that data between scaffold rods can be measured more safely and conveniently, the risks that traditional manual detection is low