| GISdevelopment.net ---> AARS ---> ACRS 2000 ---> Poster Session 1 |
A Photogrammetric Evaluation Of An Aps Camera
Tian-Yuan SHIH and Chien-Bin KUNG
Department of Civil Engineering
National Chiao-Tung University
1001 Ta-Hsueh Road, Hsin-Chu, Taiwan
E-mail:tyshih@cc.nctu.edu.tw
Tian-Yuan SHIH and Chien-Bin KUNG
Department of Civil Engineering
National Chiao-Tung University
1001 Ta-Hsueh Road, Hsin-Chu, Taiwan
E-mail:tyshih@cc.nctu.edu.tw
Keywords: Resolution, Calibration
Abstract
APS (Advanced Photographic System) is a new camera format recently introduced to the market. The APS formatted camera's photogrammetric potential is tested in this study with practical experiments. The photographic resolution is first evaluated with the USAF optical test pattern-resolving power chart. Then a test field is used for the evaluation of the metric properties. After the photographic development, the photographs are scanned and the image coordinates are then measured in a digital environment. In comparison to Wild P32 metric camera, significantly larger distortions were found with the tested Kodak Advantix 2100 Auto camera.
1. Introduction
The Advanced Photographic System (APS) is a new commercial photographic system developed by the alliance of Cannon, Kodak, Fuji, Minolta, and Nikon. The objective of APS development is to provide a more reliable commercial photographic system with better quality. Although APS is not expected to replace the commonly used 135 format camera completely, it is hoped that the APS format camera will eventually share a significant portion of the market.
The effective exposure area of traditional 135 format film is 24mm×36mm. While the film size of APS is smaller, 16.7mm×30.2mm, the image quality is not reduced. A newly developed film technique made this possible. With smaller film size, cameras can be made more compact. Usually the APS camera is 20% smaller than its 135 format counterpart. This may not be an advantage for close-range photogrammetric applications, because the photo scale is smaller for the same object distance and focal length.
There are three photo formats available in the APS system. As listed in Table 1, each photo format has a different aspect ratio. Accordingly, each photo format has different magnification scales. In this study, H format is selected.
Table 1: APS Photo Format Specification
A low-end APS camera, Kodak Advantix 2100 AUTO, is used in this study. The focal length is 25 mm with maximum aperture f/5.4. Because the APS film is designed that it will not to be accessed by the user, the photo coordinate measurements are performed with the developed photo, not the film.
2. The Resolution
LPM (lines per millimeter) is frequently applied for measuring the resolution of an imaging system. In this study, a printed standard chart is posted onto a vertical wall as the reference for evaluation. In this chart, there are groups of bars with different widths and four different colors, namely, black, red, blue, and yellow. The developed paper prints are scanned with 300, 600, 1000, and 1500dpi with a desktop flatbed scanner. The evaluation is performed on the screen of a personal computer. When the three lines cannot be differentiated, the LPM number is obtained from the look-up table printed on the reference chart. Then, the largest LPM value among all groups is taken to compute the LPM value of the photo.
LPM photo = LPM chart×(D-f0)/f0 (4-1)
Where D is the distance between the camera and the chart; f0 is the focal length of the camera.
Table 2: The Resolution of Kodak Advantix 2100 AUTO I (300dpi)
Table 3: The Resolution of Kodak Advantix 2100 AUTO II (600dpi)
Table 4: The Resolution of Kodak Advantix 2100 AUTO III (1000dpi)
Table 5: The Resolution of Kodak Advantix 2100 AUTO IV (1500dpi)
From tables 2, 3, 4, and 5, it is observed that there are differences among the photos taken with different object distances, while the scanning resolution is the same. However, if the number is normalized with photo scale, the resolution is about the same. It is also observed that the LPM improves with higher scanning resolution. But, the LPM remains the same after certain scanning resolution. That is, after certain scanning resolutions, the resolution of the photo itself becomes the governing factor.
3. The Test Fields
3.1 The NCKU Test Field
The NCKU test field is established indoors. The targets in the field are designed for close-range applications. There are three depth levels, as shown in Figure 1. The first and the second level each consist of several hanging metal strips. The targets are adhered to the strips. The targets of the third level are directly stuck onto the wall. The three dimensional coordinates of each target are determined with both total station and photogrammetric means. In the current experiment, several photos were taken from different angles with Kodak Advantix 2100 AUTO. Four of them are selected for further measurement and analysis.
Figure 1: The NCKU (left) and NCTU (right) Test Field
3.2 The NCTU Test Field
An outdoor test field is established on the campus of National Chiao-Tung University, Hsin-Chu. A building as shown in Figure 1 is used as the object. Natural points, such as the corner of windows, are selected as targets. The object coordinates are measured with both the conventional surveying method with a total station and the photogrammetric method with a Wild P32 metric camera. In this study, four projective stations are established, from the left to right, namely, STA1, STA2, STA3, and STA4.
4. The Photogrammetric Evaluation
Based on collinearity equation, the relationship between image and object space can be described. The correction terms implemented to model the deviation between the ideal and the real optical systems, and named as the additional parameters. According to Brown (1971), the physical distortion can be described with radial, decentering, and affine distortions. The program UNBASC1 (Moniwa, 1972), which includes the interior parameters (x0, y0, c) and distortion parameters (K1, K2, K3, P1, P2, A, B) is used in this study. This program does not have the gross error detection scheme implemented. Therefore, three times the amount of the RMSE is used as the threshold for screening the residuals. In order to avoid the spreading effect of gross errors, the gross errors are removed one at a time.
4.1 The NCKU Test Field
The image coordinates are measured with WINDIG program on an Intel-based personal computer. Each point is measured three times and the root mean square errors (RMSE) are listed in Table 6.
Table 6: The Measuring Repeatibility of Image Coordinates, NCKU Test Field
From Table 6, the RMSEs of the image coordinate measurement are all less than one pixel (about 0.0847mm). With the space resection, some gross errors are found. After cross-examination (cross-examination) between photos, it is suspected that the locations of these points have been displaced after determination of reference coordinates. The RMSE values also change with additional parameter sets. When all distortions are considered, the residuals become the smallest.
In the space intersection stage, different stereopairs result in different RMSE values (Table 7, 8). Because stereopair 3-7 has the longest base length, and stereopair 2-3 has the shortest object distance, these two pairs have better base/height ratio, and, therefore, have better accuracy. Meanwhile, the farther the points are, the worse the result. The points on the wall have the lowest accuracy. This is because the larger the object distance, the smaller the photo scale, and then the larger the measurement error. The farther points have relatively worse geometry. That is, the larger the base/height ratio.
Table 7: The RMSE of Image Coordinates after Adjustment with UNBASC1, NCKU Test Field, Photo 2
Table 8: Space Intersection Accuracy with UNBASC1, NCKU Test Field
(With all three distortion models; in mm)
4.2 The NCTU Test Field
Four photos are taken for the NCTU test field, namely 21, 22, 23, and 24. The photo scale is about 1/500. Scanned with 300 dpi and measured with WINDIG three times, the RMSE is about 0.03 mm for both x and y. The intersection accuracy is listed in Table 9.
Table 9: Space Intersection with Photo 21 and 24, NCTU, UNBASC1, 300 dpi
(in mm) (B/D ~ 0.24)
5. Concluding Remarks
Abstract
APS (Advanced Photographic System) is a new camera format recently introduced to the market. The APS formatted camera's photogrammetric potential is tested in this study with practical experiments. The photographic resolution is first evaluated with the USAF optical test pattern-resolving power chart. Then a test field is used for the evaluation of the metric properties. After the photographic development, the photographs are scanned and the image coordinates are then measured in a digital environment. In comparison to Wild P32 metric camera, significantly larger distortions were found with the tested Kodak Advantix 2100 Auto camera.
1. Introduction
The Advanced Photographic System (APS) is a new commercial photographic system developed by the alliance of Cannon, Kodak, Fuji, Minolta, and Nikon. The objective of APS development is to provide a more reliable commercial photographic system with better quality. Although APS is not expected to replace the commonly used 135 format camera completely, it is hoped that the APS format camera will eventually share a significant portion of the market.
The effective exposure area of traditional 135 format film is 24mm×36mm. While the film size of APS is smaller, 16.7mm×30.2mm, the image quality is not reduced. A newly developed film technique made this possible. With smaller film size, cameras can be made more compact. Usually the APS camera is 20% smaller than its 135 format counterpart. This may not be an advantage for close-range photogrammetric applications, because the photo scale is smaller for the same object distance and focal length.
There are three photo formats available in the APS system. As listed in Table 1, each photo format has a different aspect ratio. Accordingly, each photo format has different magnification scales. In this study, H format is selected.
| Name | Aspect Ratio | Format (mm) | Magnification |
| C | 2:3 | 4×6 (102×152) | 6.1 |
| H | 9:16 | 4×7 (102×178) | 6.1 |
| P | 1:3 | 4×12 (102×305) | 10.1 |
A low-end APS camera, Kodak Advantix 2100 AUTO, is used in this study. The focal length is 25 mm with maximum aperture f/5.4. Because the APS film is designed that it will not to be accessed by the user, the photo coordinate measurements are performed with the developed photo, not the film.
2. The Resolution
LPM (lines per millimeter) is frequently applied for measuring the resolution of an imaging system. In this study, a printed standard chart is posted onto a vertical wall as the reference for evaluation. In this chart, there are groups of bars with different widths and four different colors, namely, black, red, blue, and yellow. The developed paper prints are scanned with 300, 600, 1000, and 1500dpi with a desktop flatbed scanner. The evaluation is performed on the screen of a personal computer. When the three lines cannot be differentiated, the LPM number is obtained from the look-up table printed on the reference chart. Then, the largest LPM value among all groups is taken to compute the LPM value of the photo.
Where D is the distance between the camera and the chart; f0 is the focal length of the camera.
| Distance (cm) | LPM chart | LPM photo | Photo Scale |
| 100 | 0.891 | 34.7 | 1:6.20 |
| 120 | 0.707 | 33.2 | 1:7.37 |
| 150 | 0.629 | 37.1 | 1:9.27 |
| 180 | 0.500 | 35.5 | 1:11.11 |
| 240 | 0.354 | 33.6 | 1:14.51 |
Table 3: The Resolution of Kodak Advantix 2100 AUTO II (600dpi)
| Distance (cm) | LPM chart | LPM photo | Photo Scale |
| 100 | 1.000 | 39.0 | 1:6.20 |
| 120 | 0.794 | 37.3 | 1:7.37 |
| 150 | 0.707 | 41.7 | 1:9.27 |
| 180 | 0.561 | 39.8 | 1:11.11 |
| 240 | 0.397 | 37.7 | 1:14.51 |
Table 4: The Resolution of Kodak Advantix 2100 AUTO III (1000dpi)
| Distance (cm) | LPM chart | LPM photo | Photo Scale |
| 100 | 1.000 | 39.0 | 1:6.20 |
| 120 | 0.891 | 41.9 | 1:7.37 |
| 150 | 0.707 | 41.7 | 1:9.27 |
| 180 | 0.629 | 44.7 | 1:11.11 |
| 240 | 0.445 | 42.3 | 1:14.51 |
Table 5: The Resolution of Kodak Advantix 2100 AUTO IV (1500dpi)
| Distance (cm) | LPM chart | LPM photo | Photo Scale |
| 100 | 1.000 | 39.0 | 1:6.20 |
| 120 | 0.891 | 41.9 | 1:7.37 |
| 150 | 0.707 | 41.7 | 1:9.27 |
| 180 | 0.629 | 44.7 | 1:11.11 |
| 240 | 0.445 | 42.3 | 1:14.51 |
3. The Test Fields
3.1 The NCKU Test Field
The NCKU test field is established indoors. The targets in the field are designed for close-range applications. There are three depth levels, as shown in Figure 1. The first and the second level each consist of several hanging metal strips. The targets are adhered to the strips. The targets of the third level are directly stuck onto the wall. The three dimensional coordinates of each target are determined with both total station and photogrammetric means. In the current experiment, several photos were taken from different angles with Kodak Advantix 2100 AUTO. Four of them are selected for further measurement and analysis.
 |
 |
Figure 1: The NCKU (left) and NCTU (right) Test Field
3.2 The NCTU Test Field
An outdoor test field is established on the campus of National Chiao-Tung University, Hsin-Chu. A building as shown in Figure 1 is used as the object. Natural points, such as the corner of windows, are selected as targets. The object coordinates are measured with both the conventional surveying method with a total station and the photogrammetric method with a Wild P32 metric camera. In this study, four projective stations are established, from the left to right, namely, STA1, STA2, STA3, and STA4.
4. The Photogrammetric Evaluation
Based on collinearity equation, the relationship between image and object space can be described. The correction terms implemented to model the deviation between the ideal and the real optical systems, and named as the additional parameters. According to Brown (1971), the physical distortion can be described with radial, decentering, and affine distortions. The program UNBASC1 (Moniwa, 1972), which includes the interior parameters (x0, y0, c) and distortion parameters (K1, K2, K3, P1, P2, A, B) is used in this study. This program does not have the gross error detection scheme implemented. Therefore, three times the amount of the RMSE is used as the threshold for screening the residuals. In order to avoid the spreading effect of gross errors, the gross errors are removed one at a time.
4.1 The NCKU Test Field
The image coordinates are measured with WINDIG program on an Intel-based personal computer. Each point is measured three times and the root mean square errors (RMSE) are listed in Table 6.
| Photo | # of repetitions | # of Points | RMSE x | RMSE y | ||
| mm | pixel | mm | pixel | |||
| 2 | 3 | 82 | 0.0369 | 0.436 | 0.0495 | 0.585 |
| 3 | 3 | 84 | 0.0326 | 0.385 | 0.0373 | 0.441 |
| 5 | 3 | 115 | 0.0304 | 0.360 | 0.0325 | 0.384 |
| 7 | 3 | 106 | 0.0283 | 0.334 | 0.0266 | 0.314 |
| Sum | 3 | 387 | 0.0318 | 0.375 | 0.0364 | 0.430 |
In the space intersection stage, different stereopairs result in different RMSE values (Table 7, 8). Because stereopair 3-7 has the longest base length, and stereopair 2-3 has the shortest object distance, these two pairs have better base/height ratio, and, therefore, have better accuracy. Meanwhile, the farther the points are, the worse the result. The points on the wall have the lowest accuracy. This is because the larger the object distance, the smaller the photo scale, and then the larger the measurement error. The farther points have relatively worse geometry. That is, the larger the base/height ratio.
| Additional Parameters | # of control points | # of gross errors | Image coor. RMSE(mm) x y | |
| None | 57 | 1 | 0.077 | 0.065 |
| Radial | 56 | 2 | 0.062 | 0.062 |
| Decentering | 57 | 1 | 0.075 | 0.063 |
| Film | 57 | 1 | 0.062 | 0.053 |
| All | 56 | 2 | 0.039 | 0.047 |
Table 8: Space Intersection Accuracy with UNBASC1, NCKU Test Field
(With all three distortion models; in mm)
| Photo Pair | # of Control Pts | # of Check Pts | # of Gross Error | B/DValue | Control Pts RMSE | Check Pts RMSE | ||||
| X | Y | Z | X | Y | Z | |||||
| 2 and 3 | 42 | 7 | 1 | 0.28 | 1.0 | 1.0 | 4.0 | 1.1 | 1.0 | 3.0 |
| 2 and 5 | 43 | 8 | 1 | 0.14 | 1.0 | 1.0 | 7.0 | 1.4 | 2.7 | 10.0 |
| 2 and 7 | 21 | 7 | 0 | 0.15 | 4.0 | 2.0 | 13.0 | 3.0 | 2.0 | 14.0 |
| 3 and 5 | 53 | 8 | 0 | 0.10 | 5.0 | 3.0 | 19.0 | 1.1 | 1.2 | 10.4 |
| 3 and 7 | 26 | 7 | 0 | 0.38 | 1.0 | 1.0 | 4.0 | 0.7 | 1.7 | 3.3 |
| 5 and 7 | 53 | 7 | 1 | 0.25 | 1.0 | 1.0 | 5.0 | 0.7 | 0.7 | 3.2 |
4.2 The NCTU Test Field
Four photos are taken for the NCTU test field, namely 21, 22, 23, and 24. The photo scale is about 1/500. Scanned with 300 dpi and measured with WINDIG three times, the RMSE is about 0.03 mm for both x and y. The intersection accuracy is listed in Table 9.
(in mm) (B/D ~ 0.24)
| Additional Parameters | # of Control Points | # of Check Points | # of Gross Errors | RMSE on Control Points | RMSE on Check Points | ||||
| X | Y | Z | X | Y | Z | ||||
| None | 46 | 28 | 0 | 67 | 79 | 172 | 83 | 84 | 193 |
| Radial | 44 | 28 | 2 | 40 | 60 | 137 | 73 | 82 | 180 |
| Decentering | 45 | 28 | 1 | 63 | 63 | 148 | 83 | 69 | 178 |
| Film | 46 | 28 | 0 | 60 | 59 | 155 | 79 | 69 | 176 |
| All | 45 | 28 | 1 | 33 | 43 | 124 | 53 | 54 | 144 |
- Regarding the digitization of image coordinates, the scanning/digitizing approach proved to be a good replacement for conventional comparator. The scanning resolution of 450 dpi provides slightly smaller RMSE than 300 dpi. There is also an improvement in the object coordinate determination by the photogrammetric process. The image coordinates are digitized on a Wild BC-3 analytical plotter with the original negative. It is found that the additional parameters are more effective for the enlargement/scanning approach, than for the direct measurement with the negative. Apparently, additional distortions are introduced in the enlargement/scanning procedure.
- Based on the simulation, it can be concluded that the longer the focal length, the larger the photo scale, and in turn, the better the geometry and accuracy.
- Comparing the simulation with the experiments, simulation concluded better accuracy. This is understood by the fact that only random errors are considered in the simulation; while the real case suffers with the systematic errors.
- Comparing the results from APS and Wild P32, P32 has longer focal length and larger image format. The lens distortion with P32 is smaller as well. Although a 6.1 times enlargement is performed with the APS photo before the scanning, the Wild P32 still provides better accuracy.
- When the photo scale is about 1/500, the relative accuracy from APS photos can reach 1/500 of the average object distance. For the current study, the base/height ratio is limited by the surroundings of the test fields. Accuracy may be further improved if the geometry can be improved.
- Brown, D.C., 1971. Close-Range Camera Calibration, Photogrammetric Engineering, 37:855.
- Edmund, 1978. USAF 1951 Resolving Power Chart, Edmund Scientific Co., Barrington, New Jersey, U.S.A.
- Moniwa, H., 1972. Analytical Camera Calibration for Close-Range Photogrammetry, MscE thesis, Department of Surveying Engineering, Univ. of New Brunswick.
- Wolf, P.R., 1983. Elements of Photogrammetry, 2nd Ed., McGraw-Hill