The invention discloses a laser metal additive manufacturing fusion depth prediction system based on deep learning and transfer learning, which comprises a printing workbench, an image acquisition device, a thermal imager, a man-machine interaction device, a display and a host; the image acquisition device, the thermal imager, the man-machine interaction device and the display are electrically connected with the host. Molten pool images and temperature images are continuously collected in a certain time sequence; firstly, normalization processing is carried out on an effective molten pool image and an effective temperature image; the parameters of the picture size and the pixel size of the molten pool image are kept consistent, other irrelevant features are eliminated during training of the deep learning convolutional neural network model, only key features are trained, and the system has the advantage that the training efficiency of the deep learning convolutional neural network modelis improved; and the fusion depth is predicted by adopting a deep learning convolutional neural network model, so that the precision of parameters can be effectively improved.