Building outer wall hollowing leakage anomaly detection and positioning method

The invention relates to the field of artificial intelligence, in particular to a building outer wall hollowing leakage anomaly detection and positioning method, which comprises the following steps: acquiring an infrared grey-scale map of a building outer wall; according to the gray value of each pixel point in the gray-scale map, determining the pixel points of the serious hollowing area and the remaining pixel points; calculating the abnormal degree of each residual pixel point according to the gray average value of each residual pixel point and the neighborhood pixel points thereof; performing sliding window traversal by taking each residual pixel point as a central point, and obtaining the probability that each residual pixel point is a pixel point in a hollow drum region by utilizing the abnormal degree and gradient direction of each sliding window central point and the gradient direction of the pixel point in the serious hollow drum region of the sliding window; and selecting the pixel points in the ser