Метод визначення швидкості мінометного боєприпасу на основі комп'ютерного зору

The study proposes a method for determining the speed of a mortar shell based on computer vision. The method relies on video recording of the mortar shell's flight at 60 frames per second and utilizes image analysis algorithms to identify the object, determine its spatial coordinates, and calculate its speed. The proposed approach does not require specialized equipment, as it uses only a modern smartphone camera. This makes it accessible for field applications without the need for complex technologies or expensive optical systems. The algorithm involves user interaction with the program in a frame-by-frame mode to accurately determine the position of the mortar munition. By using scale conversion from pixel coordinates to the metric system, the program enables the calculation of the distance traveled between consecutive frames and the computation of the object's average speed. This process takes into account the timestamp of each frame, obtained based on the video recording frequency. Field measurements have shown that using video recordings at 60 frames per second provides optimal accuracy in determining the speed of a mortar munition's shot. The advantages of the method are its accessibility and ease of implementation, allowing it to be used in field conditions without the need for complex hardware solutions. Unlike traditional radar systems or high-speed cameras, this method enables the acquisition of accurate kinematic parameters of the munition using only a mobile device and computer vision algorithms. The obtained results confirm the effectiveness of the proposed method in real combat or testing conditions. The implementation of this technology will allow artillery units to more accurately assess the speed of munitions, which is crucial for calculating flight range and enhancing fire efficiency. Future research prospects include automating the process of munition detection using artificial intelligence, which will reduce the influence of human factors on measurement accuracy. It is also possible to expand the method for analyzing other types of projectiles and fast-moving objects, making it versatile for both military and scientific applications.