Suitability Assessment of A4 Format Desktop Publishing Scanners
Wan Mohd Naim b. Wan Mohd.
Department of Surveying Science and Geomatics
MARA Institute of Technology, Shah Alam, Selangor, Malaysia
Tel: 03-5564423
Email : naim@sug.engr.itm.edu.my
R. P. Kirby
Department of Geography
University of Edinburgh, U. K.
Abstract
High quality scanners are needed to convert the existing hardcopy aerial photographs to digital format. Although there are many types of photogrammetric scanners available in the market today, their price is still quite high. In contrast, A4 format desktop publishing (DTP) scanners are easily accessible to the general public because of their low cost. One of the problems associated with this type of scanner is the insufficient geometric accuracy caused by the mechanical instabilities, large lens distortions and imperfections of other mechanical parts such as mirrors or filters. This papers results of the geometric accuracy assessment of four different types of A4 format DTP scanners.
Introduction
Photogrammetry has been widely recognized as an important method of acquiring digital topographic information for GIS. Unlike the traditional photogrammetric method where by hardcopy aerial photographs (i. e. paper prints or diapositive) are needed, digital photogrammetry can accept data from any other sources such as stereo SPOT satellite imagines, images from CCD camera and scanned aerial photographs. In a Developing Country like Malaysia, where stereo SPOT satellite are not widely available users have to rely on existing aerial photographs s the main data source to a digital photogrammetric system.
Scanned aerial photographs are an integral part of digital photogrammetry from which it I possible to produce ortho-rectified images, ortho-images ,digital elevation models (DEM), map overlays and other products which can be used in GIS. To enable high quality products to be generated, it is essential to have geometrically and radiometrically correct images. Although high quality scanners such as the Zeiss/Intergraph Photogrammetric application, their price is quite high. On the other hand, DTP scanners are cheap general purpose scanners for converting hardcopy drawings or photographs into digital format. These types of scanners can either be in A3 or A4 format.
Carstensen and Cambell (1991) have argues that, although these scanners are not designed to be used for photogrammetric applications, they have greater potential than their manufacturers suggest. These scanners have excellent spatial resolution, a relatively low, though adequate and improving radiometric resolution, and competitive prices. According to Baltsavias (1994), some of their components like sensors, electronics, computer platforms, software, and characteristics like radiometric performance and speed are equivalent or better than those of expensive photogrammetric scanners.
Although many earliest studies (Baltsavias, 1994a; Finch and Miller, 1994; Gagnon and Argnard, 1992; Sarjakoski, 1992) have identified the potential of some of the A3 format scanners such as the Sharp JX -610 and Agfa Horizon Plus for digital photogrammetric applications, the price of acquiring A3 scanners is still high for small organizations. A4 format scanners are the cheapest, and seem to be feasible alternative for these organizations.
Assessments of A4 format DTP scanners
Although A4 scanners are cheap, the user community is widely unaware of their limitations. The issues relating to the accuracy need to be assessed and publicized. Among the issues and questions that need to be addressed are:
- What is the pattern and magnitude of distortions of images scanned in A4 format DTP scanners?
- What is the effect of scanning resolution on the pattern and magnitude of image distortion?
- What is the best mathematical formulae to be used to model and rectify distortions of images scanned in A4 format DTP scanners?
- What are the optimum number and arrangement of the image control points required to rectify image distortions?
- How repeatable are scanner measurements over time?
Only results from tests to address issues 1 and 3 will be given.
Distortion pattern and the magnitude of distortion of image scanned in DTP scanners
The main aim of this test was to determine the distortion pattern and magnitude of image distortions scanned in five DTP scanners. Four different types of DTP scanners were used, that is Apple Colour OneScanner, HP Scanjet IIc scanner, Epson GT-9000 scanner and HP Scanjet 4C (two scanners of this type were used). The choice of these scanners was based strictly on their availability. To quantify the errors, a calibrated grid plate (used to calibrate stereoplotting machines) with a regular 1 cm grid was used. A total of 47 points (25 control and 22 check points) were measured.
