|
lensfun
0.3.2.0
|
These structures and functions allow to define and examine the properties of a lens. More...
Classes | |
| struct | lfLensCalibDistortion |
| Lens distortion calibration data. More... | |
| struct | lfLensCalibTCA |
| Laterlal chromatic aberrations calibration data. More... | |
| struct | lfLensCalibVignetting |
| Lens vignetting calibration data. More... | |
| struct | lfLensCalibCrop |
| Struct to save image crop, which can depend on the focal length. More... | |
| struct | lfLensCalibFov |
| Struct to save calibrated field of view, which can depends on the focal length (DEPRECATED) More... | |
| struct | lfLensCalibRealFocal |
| Struct to save real focal length, which can depends on the (nominal) focal length. More... | |
| struct | lfParameter |
| This structure describes a single parameter for some lens model. More... | |
| struct | lfLens |
| Lens data. Unknown fields are set to NULL or 0. More... | |
These structures and functions allow to define and examine the properties of a lens.
| enum lfCropMode |
| enum lfDistortionModel |
The Lensfun library implements several lens distortion models. This enum lists them.
Distortion usually heavily depends on the focal length, but does not depend on the aperture. In the following, \(r_d\) refers to the distorted radius (normalised distance to image center), and \(r_u\) refers to the undistorted, corrected radius. See section How it is really done for further information.
For a popular explanation of lens distortion see http://www.vanwalree.com/optics/distortion.html
| Enumerator | |
|---|---|
| LF_DIST_MODEL_NONE |
Distortion parameters are unknown. |
| LF_DIST_MODEL_POLY3 |
3rd order polynomial model, which is a subset of the PTLens model. \[r_d = r_u \cdot (1 - k_1 + k_1 r_u^2)\] The corresponding XML attribute is called “k1”. It defaults to 0. |
| LF_DIST_MODEL_POLY5 |
5th order polynomial model. \[r_d = r_u \cdot (1 + k_1 r_u^2 + k_2 r_u^4)\] The corresponding XML attributes are called “k1” and “k2”. They default to 0. Ref: http://www.imatest.com/docs/distortion.html |
| LF_DIST_MODEL_PTLENS |
PTLens model, which is also used by Hugin. \[r_d = r_u \cdot (a r_u^3 + b r_u^2 + c r_u + 1 - a - b - c)\] The corresponding XML attributes are called “a”, “b”, and “c”. They default to 0. |
| LF_DIST_MODEL_ACM |
Adobe Camera Model. The coordinate system is different here. Everything is measured in units of the focal length of the lens. \[\begin{aligned} x_d &= x_u (1 + k_1 r_u^2 + k_2 r_u^4 + k_3 r_u^6) + 2x_u(k_4y_u + k_5x_u) + k_5 r_u^2, \\ y_d &= y_u (1 + k_1 r_u^2 + k_2 r_u^4 + k_3 r_u^6) + 2y_u(k_4y_u + k_5x_u) + k_4 r_u^2, \end{aligned} \] with \(r_u^2 = x_u^2 + y_u^2\). The corresponding XML attributes are called “k1”, “k2”, “k3”, “k4”, and “k5”. They default to 0. Note that Adobe's fisheye model is also covered by this by setting the lens type to "fisheye" and using only the \(k_1\) and \(k_2\) parameters. Ref: http://download.macromedia.com/pub/labs/lensprofile_creator/lensprofile_creator_cameramodel.pdf |
| enum lfLensType |
Lens type. See Change of projection for further information.
| enum lfTCAModel |
The Lensfun library supports several models for lens lateral chromatic aberrations (also called transversal chromatic aberrations, TCA).
TCAs depend on focal length, but does not depend of the aperture. In the following, \(r_d\) refers to the distorted radius (normalised distance to image center), and \(r_u\) refers to the undistorted, corrected radius. See section How it is really done for further information.
For a popular explanation of chromatic aberrations see http://www.vanwalree.com/optics/chromatic.html
| Enumerator | |
|---|---|
| LF_TCA_MODEL_NONE |
No TCA correction data is known. |
| LF_TCA_MODEL_LINEAR |
Linear lateral chromatic aberrations model. \[\begin{aligned} r_{d,R} &= r_{u,R} k_R \\ r_{d,B} &= r_{u,B} k_B \end{aligned}\] The corresponding XML attributes are called “kr” and “kb”. They default to 1. Ref: http://cipa.icomos.org/fileadmin/template/doc/TURIN/403.pdf |
| LF_TCA_MODEL_POLY3 |
Third order polynomial. \[\begin{aligned} r_{d,R} &= r_{u,R} \cdot (b_R r_{u,R}^2 + c_R r_{u,R} + v_R) \\ r_{d,B} &= r_{u,B} \cdot (b_B r_{u,B}^2 + c_B r_{u,B} + v_B) \end{aligned}\] The corresponding XML attributes are called “br”, “cr”, “vr”, “bb”, “cb”, and “vb”. vr and vb default to 1, the rest to 0. Ref: http://wiki.panotools.org/Tca_correct |
| LF_TCA_MODEL_ACM |
Adobe camera model for TCA. The coordinate system is different here. Everything is measured in units of the focal length of the lens. \[\begin{aligned} x_{d,R} &= \alpha_0 \left(\left(1 + \alpha_1 r_{u,R}^2 + \alpha_2 r_{u,R}^4 + \alpha_3 r_{u,R}^6\right) x_{u,R} + 2(\alpha_4 y_{u,R} + \alpha_5 x_{u,R}) x_{u,R} + \alpha_5 r_{u,R}^2\right), \\ y_{d,R} &= \alpha_0 \left(\left(1 + \alpha_1 r_{u,R}^2 + \alpha_2 r_{u,R}^4 + \alpha_3 r_{u,R}^6\right) y_{u,R} + 2(\alpha_4 y_{u,R} + \alpha_5 x_{u,R}) y_{u,R} + \alpha_4 r_{u,R}^2\right), \\ x_{d,B} &= \beta_0 \left(\left(1 + \beta_1 r_{u,B}^2 + \beta_2 r_{u,B}^4 + \beta_3 r_{u,B}^6\right) x_{u,B} + 2(\beta_4 y_{u,B} + \beta_5 x_{u,B}) x_{u,B} + \beta_5 r_{u,B}^2\right), \\ y_{d,B} &= \beta_0 \left(\left(1 + \beta_1 r_{u,B}^2 + \beta_2 r_{u,B}^4 + \beta_3 r_{u,B}^6\right) y_{u,B} + 2(\beta_4 y_{u,B} + \beta_5 x_{u,B}) y_{u,B} + \beta_4 r_{u,B}^2\right), \end{aligned}\] with \(r_{u,\ast} = \sqrt{x_{u,\ast}^2 + y_{u,\ast}^2}\). The corresponding XML attributes are called “alpha0” to “alpha5”, and “beta0” to “beta5”. alpha0 and beta0 default to 1, the rest to 0. Ref: http://download.macromedia.com/pub/labs/lensprofile_creator/lensprofile_creator_cameramodel.pdf |
| enum lfVignettingModel |
The Lensfun library supports several models for lens vignetting correction.
We focus on optical and natural vignetting since they can be generalized for all lenses of a certain type; mechanical vignetting is out of the scope of this library.
Vignetting is dependent on focal length and aperture. There is also a slight dependence on focus distance. In the following, \(C_d\) refers to the corrected destination image pixel brightness, and \(C_s\) refers to the uncorrected source image pixel brightness.
For a popular explanation of vignetting see http://www.vanwalree.com/optics/vignetting.html
| Enumerator | |
|---|---|
| LF_VIGNETTING_MODEL_NONE |
No vignetting correction data is known. |
| LF_VIGNETTING_MODEL_PA |
Pablo D'Angelo vignetting model (which is a more general variant of the \(\cos^4\) law). \[C_d = C_s \cdot (1 + k_1 r^2 + k_2 r^4 + k_3 r^6)\] The corresponding XML attributes are called “k1”, “k2”, and “k3”. They default to 0. Ref: http://hugin.sourceforge.net/tech/ |
| LF_VIGNETTING_MODEL_ACM |
Adobe's vignetting model. It differs from D'Angelo's model only in the coordinate system. Everything is measured in units of the focal length of the lens. \[C_d = C_s \cdot (1 + \alpha_1 r^2 + \alpha_2 r^4 + \alpha_3 r^6).\] The corresponding XML attributes are called “alpha1”, “alpha2”, and “alpha3”. They default to 0. Ref: http://download.macromedia.com/pub/labs/lensprofile_creator/lensprofile_creator_cameramodel.pdf |
| const char* lf_get_crop_desc | ( | enum lfCropMode | mode, |
| const char ** | details, | ||
| const lfParameter *** | params | ||
| ) |
| const char* lf_get_distortion_model_desc | ( | enum lfDistortionModel | model, |
| const char ** | details, | ||
| const lfParameter *** | params | ||
| ) |
| const char* lf_get_lens_type_desc | ( | enum lfLensType | type, |
| const char ** | details | ||
| ) |
| const char* lf_get_tca_model_desc | ( | enum lfTCAModel | model, |
| const char ** | details, | ||
| const lfParameter *** | params | ||
| ) |
| const char* lf_get_vignetting_model_desc | ( | enum lfVignettingModel | model, |
| const char ** | details, | ||
| const lfParameter *** | params | ||
| ) |
| void lf_lens_add_calib_crop | ( | lfLens * | lens, |
| const lfLensCalibCrop * | cc | ||
| ) |
| void lf_lens_add_calib_distortion | ( | lfLens * | lens, |
| const lfLensCalibDistortion * | dc | ||
| ) |
| DEPRECATED void lf_lens_add_calib_fov | ( | lfLens * | lens, |
| const lfLensCalibFov * | cf | ||
| ) |
| void lf_lens_add_calib_real_focal | ( | lfLens * | lens, |
| const lfLensCalibRealFocal * | cf | ||
| ) |
| void lf_lens_add_calib_tca | ( | lfLens * | lens, |
| const lfLensCalibTCA * | tcac | ||
| ) |
| void lf_lens_add_calib_vignetting | ( | lfLens * | lens, |
| const lfLensCalibVignetting * | vc | ||
| ) |
Copy the data from one lfLens structure into another.
| dest | The destination object |
| source | The source object |
| void lf_lens_destroy | ( | lfLens * | lens | ) |
Destroy a lfLens object.
This is equivalent to C++ "delete lens".
| lens | The lens object to destroy. |
| void lf_lens_guess_parameters | ( | lfLens * | lens | ) |
| cbool lf_lens_interpolate_crop | ( | const lfLens * | lens, |
| float | focal, | ||
| lfLensCalibCrop * | res | ||
| ) |
| cbool lf_lens_interpolate_distortion | ( | const lfLens * | lens, |
| float | focal, | ||
| lfLensCalibDistortion * | res | ||
| ) |
| DEPRECATED cbool lf_lens_interpolate_fov | ( | const lfLens * | lens, |
| float | focal, | ||
| lfLensCalibFov * | res | ||
| ) |
| cbool lf_lens_interpolate_real_focal | ( | const lfLens * | lens, |
| float | focal, | ||
| lfLensCalibRealFocal * | res | ||
| ) |
| cbool lf_lens_interpolate_tca | ( | const lfLens * | lens, |
| float | focal, | ||
| lfLensCalibTCA * | res | ||
| ) |
| cbool lf_lens_interpolate_vignetting | ( | const lfLens * | lens, |
| float | focal, | ||
| float | aperture, | ||
| float | distance, | ||
| lfLensCalibVignetting * | res | ||
| ) |
| lfLens* lf_lens_new | ( | ) |
| DEPRECATED cbool lf_lens_remove_calib_fov | ( | lfLens * | lens, |
| int | idx | ||
| ) |
1.8.6