抄録
Recently, pixel numbers and functions of digital cameras are amazingly increasing by modern semiconductor and digital technology, and there are many low-priced digital cameras which have more than 10 mega pixels on the market in Japan. In these circumstances, digital photogrammetry using digital cameras are enormously expected in various application fields.
There is a large body of literature on the calibration of digital cameras and circle target detection, location. Target detection and location with subpixel accuracy had been investigated as a star tracker issue, and many centering algorithms have been carried out. It is widely accepted that the least squares modeling of ellipse fitting is the most accurate algorithm. However, there are still problems for efficient close range photogrammetry. These problems are reconfirmation for the target detection and location with subpixel accuracy based on real data which were taken indoors and an indicator for estimating the accuracy of normal close range photogrammetry using digital cameras.
With this motive, the empirical testing of several algorithms for target detection and location with subpixel accuracy and an indicator for estimating the accuracy are investigated in this paper using real data which were taken indoors acquired 7 digital cameras which have 7.2 mega pixels to 14.7 mega pixels.
