Results and Discussions In order to verify the accuracy of the proposed measurement method, a semi-physical simulation test is carried out in the laboratory by using a high-precision two-dimensional rotating platform to simulate the rotating device of the artillery, and a field test is carried out in a artillery factory. Both semi-physical simulation test and field test have verified the measurement accuracy and repeatability deviation of the measurement system. The semi-physical test results show that the measurement accuracy of barrel altitude angle is better than 0.09° [Fig. 5(a)], and that of direction angle is better than 0.15° [Fig. 5(b)]. The repeatability deviation of both angles is better than 0.0035° (Fig. 6). The field test results show that the measurement accuracy of the measurement system for altitude angle is better than 0.15° (Table 1), while that for direction angle is better than 0.12° (Table 1), and the repeatability deviation of both angles is better than 0.0035° (Fig. 9). The field test results are basically consistent to those of semi-physical simulation test, and the repeatability deviation is very low, which shows that the measurement method is highly robust to external disturbances such as light instability.

The artillery adjustment accuracy directly affects the hitting accuracy of the artillery, and it is a key performance parameter of the artillery. The system deviation of a fire control system before artillery equipment or after serving for a period of time is generally large, which results in a large deviation between the actual direction of the barrel and the setting direction of the system. The accuracy of artillery adjustment can be evaluated by using high-precision barrel pointing measurement, and the system deviation of the fire control system can be corrected by using pointing measurement data to improve the hitting accuracy. At present, the double theodolite method is widely used for barrel pointing measurement. The theodolite needs to acquire the horizontal angle and zenith angle data of observation points in the form of artificial sighting, and the measured data need to be manually input into the computer to calculate the barrel pointing. The low measurement efficiency and low degree of automation restrict the production efficiency of the artillery, and it is difficult to meet the rapid calibration requirement of the fire control system in future battlefield environments.

The visual measurement method features non-contact, speediness, high accuracy, and easy integration. In this paper, a new method of barrel pointing measurement based on binocular vision is proposed. With corner points of a chess board fixed on the barrel as cooperative marking points, images taken by two cameras when the barrel is at zero position and measuring position are automatically analyzed by image processing, and the pixel coordinates of cooperative marking points in the images taken by two cameras are calculated and used to analyze barrel rotation. It is proposed to decompose the rotation of the barrel into two revolving movements, with only the direction angle and altitude angle changing, respectively. On this basis, measurement equations are established, and the direction angle and altitude angle are decoupled. The direction quantity of the plumb line in the camera coordinate system is obtained by plumb line measurement, and the relationship between the camera coordinate system and the geodetic coordinate system is established, with the initial value of the barrel in the geodetic coordinate system obtained. Then, LM (Levenberg-Marquardt) algorithm is used to optimize the barrel pointing, and the final result of barrel pointing is obtained.

In this paper, an artillery barrel pointing measurement system based on binocular vision is built, and measurement equations are derived. The measurement system software is developed, and the functions of system calibration and measurement are integrated to realize automatic measurement. The introduction of plumb line in the measurement method brings the advantage of low operation difficulty compared with the existing visual measurement methods and improves the convenience of visual measurement methods in the application of barrel pointing measurement. Barrel pointing measurement can be accomplished when the cooperative marking points are arranged arbitrarily without installation error calibration. The semi-physical test and field test results prove that the method presented in this paper has the advantages of high measurement accuracy, high robustness, and full-automation. It provides a new scheme for the pointing measurement of artillery barrel, which can realize the automatic measurement of artillery barrel pointing in future battlefield environments and has broad application prospects.