Research on Improved A * Algorithm for Camera Path Planning in VR Movie Shooting

The use of A * algorithm for camera path planning is crucial for introducing scene information in virtual reality (VR) movie shooting. In order to solve the problems of long redundant paths, slow pathfinding speed, multiple turning points and large turning angles in the traditional A
^*
algorithm during the pathfinding process, resulting in abrupt and unnatural transitions in the captured movie images, this paper improves the traditional A
^*
algorithm. Firstly, by calculating the ratio of the standard deviation from the current point to the target point and the standard deviation from the starting point to the target point, the weights of the heuristic function of the A
^*
algorithm are adaptively adjusted to achieve a balance between path length and path quality during the pathfinding process; Then, by changing the selection method of the extended sub nodes in the algorithm, the number of inflection points in the path is reduced without increasing the path length; Finally, the A * algorithm is combined with a cubic B-spline curve, and the path planned by the A * algorithm is smoothed using the local smoothness characteristics of B-splines. The experimental results show that the improved A * algorithm has faster path finding speed, fewer redundant inflection points in the path, smoother path, and the movie captured by the camera according to the planned path is more coherent and smooth compared to the traditional A * algorithm, while ensuring comprehensive scene information.