To solve the problems of small area, slow calculation speed and low precision in the projection transformation algorithm of raster data, the idea of using Kriging interpolation approximate grid algorithm for raster data projection transformation is proposed. Through the experimental results of point-to-point transformation and Kriging interpolation approximate grid algorithm transformation under the same conditions, it can be seen that for different projection types under the same limit conditions, Kriging interpolation approximate grid algorithm can ensure that the raster data projection error is always within the given projection limit. With the same number of points, the Kriging interpolation approximate grid algorithm is faster and more efficient than the point-to-point projection algorithm. Under the same pixel condition, the Kriging interpolation approximate grid algorithm is faster, more accurate and more effective than the point-to-point projection algorithm.