The invention relates to the technical field of simulation control and graphic processing, and provides a virtual minimally invasive abdominal aorta vascular surgery collision detection method comprising the following steps of 1, abdominal aorta vascular three-dimensional modeling; 2, abdominal aorta vascular deformation model construction; 3, rough collision detection; and 4, precise collision detection. The problems of a classic bounding box method that intersection testing is complex and efficiency is low can be solved by utilizing rapid collision detection of an external layer AABB bounding box so that real-time performance of the system can be enhanced. Precise collision detection is performed by utilizing the characteristic of tightness of an internal layer K-DOPs bounding box so that high accuracy of the detection result is guaranteed.