Recently, the number of pixels of amateur digital cameras is amazingly increasing by modern semiconductor and digital technology. Nevertheless the highest pixel of amateur digital camera was 0.8 mega pixels in the 1996, 5.24 mega pixels was achieved in the 2001 and 3 mega pixels amateur digital cameras became the most popular camera on the market in Japan. In these circumstances, it is expected that the high resolution amateur digital cameras will become useful tool in various photogrammetric fields. With this objective, the authors have been concentrating on developing a low-cost photogrammetric system using a digital amateur camera. This paper reports performance evaluation of 3 mega pixels amateur digital camera, a low-cost photogrammetric system, and the software“3DiVisoin”which was designed to perform convenient 3D measurements using amateur digital camera with key words: 3Dimension, Digital image and Visualization. Furthermore, 3D modeling of historical structure which was designated as important cultural assets using the 3DiVision was introduced.
Accuracy of image registration is severely affected by that of corresponding control point (CCP) selection in remote sensing or GIS (Geographic Information System) . In this paper, a new automated system for CCP candidate selection from target images is proposed. In the system, first, grayscale image within a quasicircular field of view (FOV) is transformed into binary one after intensity modification for the several extreme intensity pixels. Next, the binary image is transformed into a rotation invariant intermediate representation. Finally, the system determines whether the central pixel of the FOV is appropriate as the CCP by using well-trained 3-layer feedforward artificial neuralnet. Pseudo Zernike moments are employed as the intermediate representation. Consequently, without selection accuracy deterioration, we achieve quite fewer training patterns, shorter training time, and higher noise tolerance in comparison with conventional neuralnet-based systems.
Landsat-5 had acquired 12, 328 data in total over Japan (42 scenes) during 17 working years.That means 293.5 data in average for one scene were registered in that period. However, examining Cloud Coverage Ratio (CCR) in whole data, clear images were very limited because optical sensor was hindered by cloud. The highest sites of clear sky frequency through year (CCR≤20; CCR less than 20%) were Kanto (WRS 107-35), fo11owed by northeastern Hokkaido (WRS 106-29), 13.71% and 11.42%, respectively (Fig.1) .The worst two were northern Okinawa (WRS 113-41) and southern Okinawa (WRS 113-42), 1.52% and 1.78%, respectively. Dividing Japan into 6 blocks, Kanto & Tokai block wascharacterized as high during winter seasons (12.72%), while Hokkaido & Tohoku Pacific side was high in summer seasons (6.99%) .Comparing with some arid areas in foreign countries, there was a drastic difference. CCR0 (no cloud cover) and CCR20 (1-20%) were 14.00% and 39.00% for southern Australia (WRS96-82), 16.91% and 27.94% for Inner Mongolia of China (WRS 124-29, 30), and 1.27% and 4.90% for Japan (42 scenes), respectively.Under this situation, special devices and efforts are necessary for thepractical use in agriculture and forestry fields.