Method and apparatus for automatic intrinsic camera calibration using images of a planar calibration pattern
The present disclosure relates to a high precision method, model, and apparatus for calibrating the intrinsic camera parameter values of one or more intrinsic parameters of a camera. The intrinsic parameters include focal lengths, skew, optical center, and radial distortion coefficients. The practical implementation of the solution consists of a camera capturing several images of a planar calibration pattern from different viewpoints, and a computer implemented processing pipeline. The solution does not require additional human input and is robust to misdetection of the calibration pattern in some of the images.
1. A method comprising:
for each image of a plurality of images taken by an image capture device of a calibration pattern, reconstructing a lattice graph corresponding to the calibration pattern;
determining two-and three-dimensional correspondences for the plurality of images, based on the reconstructed lattice graphs;
randomly selecting a plurality of sets of images from the plurality of images;
for each set of images in said plurality of sets of images:
determining a set of estimated values for a plurality of intrinsic parameters for the image capture device, based on the determined two-and three-dimensional correspondences for the images in the set of images, and
based on said set of estimated intrinsic parameter values, calculating a reprojection error for each of the plurality of images;
determining a measure of quality for each set of estimated intrinsic parameter values, based on the reprojection errors calculated for that set of estimated intrinsic parameter values;
selecting one of the sets of estimated intrinsic parameter values, based on the measures of quality for the sets of estimated intrinsic parameter values;
selecting a subset of images from the plurality of images with respective reprojection errors that are based on the selected set of estimated intrinsic parameter values and are lower than a reprojection error threshold; and
determining a set of re-estimated intrinsic parameter values, based on the subset of images and the two-and three-dimensional correspondences corresponding to the subset of images.
2. The method of claim 1 , wherein determining a measure of quality for each set of estimated values for the plurality of intrinsic image capture device parameters comprises determining a median of the reprojection errors corresponding to the set of estimated values.
3. The method of claim 2 , wherein the reprojection error threshold is derived from the median of the reprojection errors corresponding to the set of estimated values.
4. The method of claim 1 , further comprising updating current intrinsic parameter values of the image capture device, based on the set of re-estimated intrinsic parameter values.
5. The method of claim 1 , wherein the plurality of intrinsic parameters are selected from the group consisting of: focal length, skew, optical center, and radial distortion coefficients.
6. The method of claim 1 , wherein each lattice graph includes a plurality of nodes, each of which corresponds to at least a portion of a respective one of a plurality of features of the calibration pattern.
7. The method of claim 1 , further comprising obtaining the plurality of images from frames in a video.
8. The method of claim 1 , wherein the plurality of images is a series of still images.
9. The method of claim 1 , wherein the determining of the two-and three-dimensional correspondences for the plurality of images uses an arbitrarily selected node of each lattice graph as the origin.
10. The method of claim 1 , wherein determining a measure of quality for each set of estimated values for the plurality of intrinsic image capture device parameters comprises determining a measure of central tendency that is robust to outliers.
11. The method of claim 1 , wherein the reprojection error threshold is derived from the median of the reprojection errors corresponding to the set of estimated values.
12. A non-transitory computer readable storage medium, on which is stored computer-executable instructions that, when executed by at least one processor, cause the at least one processor to carry out the method of claim 1 .
13. A computing unit comprising:
at least one processor; and
memory storing a plurality of computer-executable instructions that, when executed by at least one processor, cause the at least one processor to carry out the method of claim 1 .