Accurate and Fast Fire Alignment Method Based on a Mono-binocular Vision System

Obtaining 3D coordinates of fire by binocular vision is beneficial for intelligent fire extinguishing. However, binocular vision technology has a small field of view (FOV) for long-distance targets, which results in the fire not appearing in the FOV of binocular cameras. To solve this problem, an automatic fire alignment system with monocular and binocular cameras is constructed, and an alignment control method, including a one-time alignment method and visual servo is proposed. First, the one-time alignment method based on the monocular camera is designed to solve the problem of time. The alignment angle independent of depth information ensures that the binocular camera can reach the vicinity of the fire in a short time with only a single camera. Then, image-based visual servo (IBVS) methods based on a monobinocular vision system are designed to improve the alignment precision of the binocular camera, and a logic controller is designed for automatically switching the vision servo control method based on different cameras. Finally, experiments are designed to compare the proposed method with other methods. Results demonstrate that the method in this paper is better than other methods in terms of both time and accuracy and contributes to improving the efficiency of fire extinguishing.