A Cluster-based Parallel Geometric Correction Algorithm for Photogrammetry

With the rapid development of data acquiring technology, the amount of image data that need to be geometrically corrected in photogrammetry becomes huge, thus processing time is becoming unbearable, especially in some applications that require real-time processing. To solve this kind of problem, the parallel geometric correction algorithm is studied, and a new parallel algorithm based on cluster is put forward. Load balanced data division strategy are analyzed and the issue of data communication is discussed. Experiments show that new algorithm can achieve good performance and has favorable expansibility.