Geo-Image Technology Research Unit
Department of Survey Engineering
chulalongkorn university, Bangkok, Thailand
Thanate Jongrugenun thanate
Chalermchon Satirapod chalermchon
Geo-Image Technology Research Unit,
Dept. of Survey Engineering, Chulalongkorn University, Bangkok, Thailand;
Dept. of Geomatics, University of Melbourne, Australia
A very attractive proposition for mapmakers in the developing world would be to have the capability of generating the metre-level accuracy of Precision imagery at the price of base-level Geo imagery. This paper addresses the means to realize this prospect through discussion of the results of experimental testing of stereo IKONOS Geo imagery
High resolution satellite imagery (HRSI) has generated much interest in the mapping and geo-information community in Thailand. To date, any mapping project using aerial photographs must seek permission from the military, the Royal Thai Survey Department (RTSD), that also controls all the processes from taking the photographs to archiving and distributing the data. Although lately control has been relaxed, the time period required before users get the data is still quite considerable. The availability of HRSI in the international market, coupled with the promise of fast data delivery, is much welcomed by users. However, for a large mapping project the price of HRSI may not be as competitive at present as aerial photograph, though this is still the subject of much debate.
There are situations in which HRSI clearly has advantages over aerial photographs. An example in the Thai context is the mapping of small remote areas, especially those close to national borders where mobilization of airplanes can be very expensive and flying airplanes is normally prohibited by the Government in these regions. Another example is the mapping of islands and off-shore natural resources such as coral reefs. There are more than 400 coral reefs in Thai waters and a substantial portion of the economy, particularly fishing and tourism, are underpinned by these resources. Again the problem here is either the high cost of mobilizing the plane to take only a few photographs or the inaccessibility of the areas, which may lie too close to border areas. As a result, there are areas where large and medium scale maps are non-existent. Using satellite imagery is arguably the only practical way to tackle the situation. But spatial resolution provided by previous generation spaceborne imaging systems such as Landsat or SPOT is too low for many mapping purposes. For example, in the case of coral reef mapping, a study by Trisirisatayawong and Samanloh (2003) that used Landsat 7 imagery to detect small shallow-water coral reefs, which are the majority type in Thailand, showed that only limited success can be achieved and higher spatial resolution imageries are required for more detailed reef mapping. HRSI exhibits significant spectral advantages over lower resolution satellite imaging systems such as Landsat and SPOT, as well as strong geometric capabilities that previously have not been available. These include much better photogrammetric stereo capability and a higher revisit rate (Li, 1998).
A large number of technical papers have demonstrated the metric capability of HRSI for large scale mapping. A user can opt for off-the-shelf image products such as 4-m accurate (CE90) Precision Plus IKONOS imagery provided by Space Imaging. The users, however, have to supply ground control points to the company. Alternatively, users can purchase less accurate but also less expensive products such as Geo or Reference and get more accurate output by applying further processing themselves. In a country of low-wages, but with a solid technological base, like Thailand, the latter approach seems to be more attractive. The gain from purchasing cheap product can easily offset the cost incurred from extra data processing.
This paper addresses the means to realize this prospect through discussion of the results of experimental testing of stereo IKONOS Geo imagery in Northern Thailand. The paper reports positional accuracy of products generated from IKONOS stereo images by applying a bias-correction technique as described below. Result from a second sensor orientation model are also shown.