Details
CAMIMPORT imports camera-calibration parameters from an XML file into an OrthoEngine project file.
About camera calibration
- Curvature of the lens
- Focal length of the lens
- Effects of perspective
The information is used to compute the interior orientation, which defines the geometry of the camera. The geometry is used to reconstruct a bundle of rays from a set of image points.
Camera-calibration report
Typically, images acquired with a photogrammetric-aerial camera include a calibration report that provides information about the camera. If images from the camera do not include a report, you can obtain one from the camera manufacturer or from a company that calibrates cameras of the type you have.
The camera-calibration report typically includes the information described in the following table.
| Parameter | Description | Required for scanned-film* | Required for digital / UAV |
|---|---|---|---|
| Camera name | Name of the camera used to acquire the imagery. | ||
| Focal length | Distance between the focal point of the lens and the film.
Note: Entering an incorrect focal length can introduce distortions in a project.
|
✔ | ✔ |
| Principal point of symmetry | Point on the image at which a ray of light that travels perpendicular to the image plane passes through the focal point of the lens and intersects the film.
With a correctly assembled camera, the principal point of symmetry is where the lines of opposing fiducial marks on an image intersect. With most cameras, however, a slight offset occurs typically. The perspective effects in the image are radial about this point. The offsets are typically specified in the camera-calibration report and are referred to as the point of symmetry or the calibrated principal point. |
||
| Radial-lens distortion | Symmetric distortion due to imperfections in the curvature of the lens during manufacturing.
Typically, errors introduced by a radial-lens distortion (ranging from approximately 1 um through 2 um) are much smaller than the scanning resolution of the image (approximately 25 um). Entering the values can significantly increase the processing time while contributing little to the final product. Values for the radial-lens distortion can be provided as ranging from R0 through R7 coefficients or in a tabular format. The equation for the lens distortion is as follows:
Where:
If you use a camera-calibration report from the U.S Geological Survey (USGS), the coefficients are given as K0, K1, K2, K3, and K4, which correspond to R1, R3, R5, and R7. K4 is discarded because it is usually zero. The coefficients may or may not appear in the camera-calibration report. It is recommended that with a digital camera you obtain the radial-distortion values. That is, radial distortion with digital cameras and their lenses tends to be greater, because their precision of manufacture tends to be inferior to that of high-end photogrammetric cameras. |
||
| Decentering distortion | Nonsymmetric distortion caused by the misalignment of the lens elements when the camera is assembled.
The decentering-distortion values can be provided ranging from P1 through P4 coefficients or in tabular format. The equation for decentering distortion is as follows:
Where:
The coefficients may or may not appear in the camera-calibration report. It is recommended that with a digital camera you obtain the decentering-distortion values. That is, decentering distortion with digital cameras and their lenses tends to be greater, because their precision of manufacture tends to be inferior to that of high-end photogrammetric cameras. |
||
| Fiducial marks | Small crosses or small v-shaped indents located precisely on each of the four corners, exactly midway along the four sides of a standard-aerial photograph, or both.
When you identify the fiducial marks in a scanned image, the fiducial marks from the camera-calibration report are used to establish a frame of image coordinates. Tip: If your scanned imagery does not have fiducial marks, you can collect them by running the CATALYST Professional AUTOFID algorithm.
Note: Images acquired with digital cameras do not have fiducial marks.
|
✔ | |
| Chip size | Physical size of the charged-coupled device (CCD) in a digital camera.
Images from digital cameras do not contain fiducial marks; therefore, the size of the CCD is used to calculate the geometry of the camera. With most cameras the sensor cells are square; however, some (especially video cameras) can have sensor cells that are rectangular. This parameter is required for images acquired with digital cameras. |
✔ | ✔ |
| Y-scale factor | Ratio between the horizontal and the vertical size of each sensor cell in digital cameras.
The y-scale factor is used when the CCD pixels are not square. Using the chip size and y-scale factor, the digital image is converted automatically to a normalized, square-image coordinate system. The image can then be processed during the computation of the math model (the bundle adjustment) in the same way as an image acquired with a scanned-film camera. |
Format of the camera-calibration file
The camera-calibration file, camera_calib.xml, is stored typically in the same folder as the input imagery. It must contain, at a minimum, the focal length of the camera used to acquire the imagery and the chip size of the charge-coupled device (CCD).
Scanned-film camera
The following example shows a proposed configuration structure for a scanned-film camera.
<CAMERA NAME="my_calib_name"> <LENS_SERIAL>my_serial</LENS_SERIAL> <CALIB_SOURCE>my_calib_source</CALIB_SOURCE> <CALIB_DATE>my_date</CALIB_DATE> <FL>90</FL> <StdDevFL>0.01</StdDevFL> <SYMMETRY PPX="0.27" PPY="0.13"/> <STDDEV PPX="0.001" PPY="0.001"/> <RLDC name="R0">0.67</RLDC> <RLDC name="R1">0</RLDC> <RLDC name="R2">0</RLDC> <RLDC name="R3">0.46</RLDC> <RLDC name="R4">0</RLDC> <RLDC name="R5">0</RLDC> <RLDC name="R6">0</RLDC> <RLDC name="R7">0</RLDC> <DDC name="P1">0</DDC> <DDC name="P2">0</DDC> <DDC name="P3">0</DDC> <DDC name="P4">0</DDC> <FILM> <EDGECORNER/> <FID_COORD name="TopLeft" FID_X="1.3" FID_Y="1.5"/> <FID_COORD name="TopMiddle" FID_X="1.5" FID_Y="1.3"/> <FID_COORD name="TopRight" FID_X="1.3" FID_Y="1.5"/> <FID_COORD name="RightMiddle" FID_X="1.5" FID_Y="1.3"/> <FID_COORD name="BottomRight" FID_X="1" FID_Y="1"/> <FID_COORD name="BottomMiddle" FID_X="1" FID_Y="1"/> <FID_COORD name="BottomLeft" FID_X="1" FID_Y="1"/> <FID_COORD name="LeftMiddle" FID_X="1" FID_Y="1"/> </FILM> </CAMERA>
Digital camera
The following example shows a proposed configuration structure for a digital camera.
<CAMERA NAME="my_calib_name"> <LENS_SERIAL>my_serial</LENS_SERIAL> <CALIB_SOURCE>my_calib_source</CALIB_SOURCE> <CALIB_DATE>my_date</CALIB_DATE> <FL>100.5</FL> <SYMMETRY PPX="1.400" PPY="0.000" /> <RLDC name="R0">0</RLDC> <RLDC name="R1">0</RLDC> <RLDC name="R2">0</RLDC> <RLDC name="R3">0</RLDC> <RLDC name="R4">0</RLDC> <RLDC name="R5">0</RLDC> <RLDC name="R6">0</RLDC> <RLDC name="R7">0</RLDC> <DDC name="P1">0</DDC> <DDC name="P2">0</DDC> <DDC name="P3">0</DDC> <DDC name="P4">0</DDC> <DIGITAL> <CHIP_SIZE WIDTH="103.860" HEIGHT="67.860" /> <YSCALE>1.000</YSCALE> </DIGITAL> </CAMERA>
Each XML tag in the camera-calibration file represents a specific parameter required for the development of an airphoto math model.
- CAMERA NAME: Name of the camera used to acquire the imagery
- FL: Focal length of the camera used to acquire the imagery
- SYMMETRY PPX PPY: Principal-point offset of the camera
- RLDC R0 - R7: Radial-lens distortion parameters of the camera
- DDC P1 - P4: Decentering-distortion parameters of the camera
- CHIP_SIZE WIDTH HEIGHT: Size, in millimeters, of the CCD used by the camera
- YSCALE: Y-scale factor of the camera used
- FILM: The following parameters affect the positioning of the fiducial marks.
- EDGE: Fiducial marks occur along the edge of the imagery.
- CORNER: Fiducial marks occur at the corners of the imagery.
- EDGECORNER: Fiducial marks occur along the edge and at the corners of the imagery.
- FID COORD: X and y coordinates of the fiducial marks, in milimeters.
If your images align along flight lines, but not between them, it is possible that the principal-point offsets are incorrect and need to be adjusted. This occurs most often if you have rotated the kappa value of your EO file. If so, your camera-calibration file will no longer be accurate. For example, if the kappa has been rotated by 90 degrees, use the information in the following table to correct the principal-point offset (PP) of the camera calibration.
| Original PP | Kappa rotation | Modified PP |
|---|---|---|
|
PP_X = x_mm |
0 degrees clockwise |
PP_X = x_mm |
|
PP_X = x_mm |
90 degrees clockwise |
PP_X = y_mm |
|
PP_X = x_mm |
270 degrees clockwise |
PP_X = -y_mm |
|
PP_X = x_mm |
180 degrees clockwise |
PP_X = -x_mm |
- When you process a small amount of data, and you notice that your images look compressed in one direction, check to see if the chip-size width and height are reversed in your camera-calibration file.
- Camera-calibration files will be the same for the same camera on the same plane and can be reused repeatedly. It is recommended that you retain them in a central, easily accessible location for future reference.
To process aerial imagery, you must have an XML file of camera-calibration information as described herein. The camera-calibration information is based on the camera-calibration report that comes with the imagery or that you obtain from the camera manufacturer. If you do not have an XML file of this information, you can create one yourself, as shown in the preceding example.
Each tag in the XML file represents a particular parameter required for the development of an airphoto math model. A document type definition (DTD) file is also required to validate the XML file so that it can be read. To help you with creating and validating your camera-calibration file (.xml), a sample file (camera_calib_sample.xml) and a sample DTD file (camera_calib.dtd) are available in the etc folder of your installation.