Camera Parameter

The (x,y,z) coordinates of points are computed based on some parameters used during imaging. Users may not be interested in these parameters, but Photo4D needs to know the values of these parameters, either given by users or computed from reference points and marks, in order to compute the point coordinates.

Each image is associated with the following camera parameters:

• Camera location Tx, Ty, Tz (in the unit of world coordinate system): They are the (x,y,z) coordinates at which the image was taken in the project's world coordinate system.
• Location range TR (in the unit of world coordinate system): TR specifies a range for Photo4D to adjust Tx, Ty, Tz. For example, if Tx = 100 and TR = 50, then Photo4D computes Tx in the range of [Tx-TR, Tx+TR] = [50, 150].
• Look-at direction Rx Ry Rz (in RAD): They indicate how the cameras are rotated along X, Y, and Z axes to look at the object. Notice that the rotation unit is in RAD, not in DEG.
• Direction range RR (in RAD): RR specifies a range for Photo4D to adjust Rx, Ry, and Rz in [Tx/y/z-TR, Tx/y/z+TR] respectively.
• Focal length F (in millimeter): F is the focal length of the camera lens.
• Focal length range FR: FR specifies a range for Photo4D to adjust F in [F/FR, F*TR].
• Distortion factor K: K models the lens distortion from the camera. It is 1.0 if there is no lens distortion, i.e., a straight line on object is always shown as a straight line in the image. However, most cameras introduce distortion, i.e., a perfectly straight line on an object appears with a slightly curved line in the image, especially when a wide-angle lens is used. K is a factor that compensates this distortion. If you do not know the K value of your camera, use its default value 1.0. Photo4D can also compute K.
• Pixel aspect ratio A: Most imaging devices produce images with A = 1.0, which is the default value. If the image device does not have 1:1 aspect ratio for pixel, this value should be set according to the formula: A = Square-root(pixel-width / pixel-height), where pixel-width and pixel-height can be found in the camera specifications. Photo4D can compute A if there are more than five reference points. However, users are encouraged to set this parameter from camera specifications, instead of letting Photo4D compute it. It is usually hard to measure a sufficient number of reference points.
• Imaging center Cc Cr (in column/row): (Cc, Cr) indicates the column/row location of imaging center in the image. By default, they are set to the middle of the image, e.g., Cc=160.0, Cr=140.0 for a 320x280 image. However, users have to set them if the image was not cropped around the center of the image before loading into Photo4D. Photo4D CANNOT compute Cc and Cr.
• Equivalent film or sensor width W (in millimeter): W is the width of the part of film (if the image was taken by camera) or sensor (if the image was taken by video camera) that covers the image. If the image is the original image developed from the film or extracted from the video, W is the width of the film or sensor. If the image was cropped, W is (original film or sensor width) * (image width) / (original image width).

Before the first computation, users need to Set Initial Camera Parameter.