Low-latitude UAV Image Processing Based On The CIPS

The low-latitude remote sensing system of Unmanned Aerial Vehicle (UVA) has beenextensively used in many fields due to its many advantages, such as flexibility, efficiency,low cost, little influence by the weather conditions, and the ability of obtaining highresolution images. Particularly, the system has enormous superiority in mapping smallareas and areas that are difficult to reach, as well as in updating the survey and mappingresults. When compared with traditional photogrammetry, however, the low-latituderemote sensing system of UVA has certain disadvantages: lack of stability in the cameraplatform; large amount of small UVA images; large obliquity; irregular overlap etc. Thesedisadvantages cause a series of difficulties in dealing with aerial Triangulation, imagematching processing; they also bring new challenges to the digital photogrammetricsystem.CIPS(Clusted Image Processing System) is a set of clustered photogrammetric processingsystem which is build in grid computing environment. It integrates the networkcommunication, computing, storage and other resources through the application of highperformance parallel computing, mass storage, and network communication technology.Therefore, it is appropriate for quick, batch processing of massive remote sensing imagesand has a good adaptability in the UAV remote sensing image processing.This article discusses the following aspects about the UAV remote sensing imageprocessing:(1) Introducing the development of UAV remote sensing and the characteristics ofUAV low-altitude photogrammetry. Both at-home and abroad research status of UAVremote sensing image processing and the UAV remote sensing image acquisition processare discussed in detailed.(2) Introducing the composition of CIPS image processing system and the functionmodules of the system, focusing on the characteristics of CIPS image processing systemand cluster parallel processing mechanism.(3) Discussing several key problems in UAV remote sensing image processing.The common algorithms to the UAV image feature extractioning, image matching, and the image fusion are compared.(4) Describing the CIPS-based UAV remote sensing image processing system andproviding possible solutions to the problems of data processing. As an experiment, two setsof UAV data are processed to test the accuracy of CIPS automatic processing and thespeed of the cluster parallel processing mode. The experiment suggests the reliability andfeasibility of fast processing of UAV image based on CIPS.