All-round panoramic monitoring method, device and system

The invention relates to a look-around panoramic monitoring method. The method comprises the following steps: obtaining a panoramic image and multi-dimensional laser point cloud data of a monitoring scene; training a neural network by adopting the training image set to obtain an image classifier, and detecting and identifying a monitoring target in the panoramic image through the image classifierto obtain a first target identification result; separating foreground points in the multi-dimensional laser point cloud data, and clustering the foreground points by using a clustering algorithm to obtain a second target recognition result; and mapping the second target identification result to the panoramic image, and fusing the first target identification result and the second target identification result by adopting a DS evidence method to obtain a fused target identification result. According to the invention, 360-degree panoramic monitoring can be achieved, and the three-dimensional structure position information