The containment system is the key device that is used to ensure safe shipping of low-temperature LNG on LNG carriers. The insulation panel is an essential component of the containment system, and its installation accuracy can directly affect the thermal insulation performance of the system. According to the requirements of the containment system’s insulation panel, a visual recognition system was developed for the installation of the insulation panel. The visual recognition system primarily consists of the recognition of the installation target and the installation surroundings. Video signals are captured by camera, subsequently, static and dynamic targets in the installation environment are identified using frame difference and convolutional neural networks. During installation, a laser visual sensor is employed to acquire point cloud images of bolts and insulation panels, then initial positioning is achieved through edge feature extraction and boundary fitting. The proposed system provides an improved visual recognition technology framework for insulation panel installation, while achieving intelligent auxiliary positioning of insulation panels.