Abstract: Effective and high-precision non-destructive detection methods for building diseases are lacking, and traditional manual detection imposes limitations on the height and area of the disease area of the building envelope facing layer. To address these issues, this study proposes a method of 3D fusion of infrared imaging and oblique photography along with the establishment of 3D model electronic archives containing building disease information. Taking a teaching building in a school as an example, this study adopts the acquisition scheme mainly based on infrared technology and supplemented by tilt photography technology and the coordinate matching method in the same position space, obtains the refined three-dimensional model electronic file containing building disease information through coordinate conversion and two kinds of heterogeneous spatial data fusion, and completes the comparative evaluation of model accuracy before and after data fusion. The results show that the fusion model obtained by this method has a high accuracy and small point error and can quickly and accurately obtain the spatial position of building diseases. The method provides a new idea for the practical application of building nondestructive testing technology and has research value and practical application significance for establishing an integrated operation and maintenance system of building information supervision, repair, and management.