X-Git-Url: http://git.maemo.org/git/?a=blobdiff_plain;ds=sidebyside;f=docs%2Fref%2Fopencvref_cv.htm;fp=docs%2Fref%2Fopencvref_cv.htm;h=da057d8fb137af78c05c9c1e6ecd2a8b15132ad3;hb=80cd7b93506cc1926882d5fd08a2c74ee9359e29;hp=0f98b1544e9aca5619b749524d0f972b58d83180;hpb=467a270adf12425827305759c0c4ea8f5b2b3854;p=opencv diff --git a/docs/ref/opencvref_cv.htm b/docs/ref/opencvref_cv.htm index 0f98b15..da057d8 100644 --- a/docs/ref/opencvref_cv.htm +++ b/docs/ref/opencvref_cv.htm @@ -32,6 +32,7 @@
CV_INTER_NN
method.
void cvSmooth( const CvArr* src, CvArr* dst, int smoothtype=CV_GAUSSIAN, - int param1=3, int param2=0, double param3=0, double param4=0 ); + int size1=3, int size2=0, double sigma1=0, double sigma2=0 );
param1
×param2
neighborhood of the pixel.
+ size1
×size2
neighborhood of the pixel.
If the neighborhood size varies from pixel to pixel, compute the sums
using integral image (cvIntegral).
-param1
×param2
neighborhood of the pixel.
+size1
×size2
neighborhood of the pixel.
param1
×param2
.
- param3
and param4
may optionally be
+ of aperture size size1
×size2
.
+ sigma1
and sigma2
may optionally be
used to specify shape of the kernel.
param1
×param1
.
+ of size size1
×size1
(i.e. only square aperture can be used).
That is, for each pixel the result is the median computed
- over param1
×param1
neighborhood.
+ over size1
×size1
neighborhood.
param1
and
- space sigma=param2
.
+ with color sigma=sigma1
and
+ spatial sigma=sigma2
. If size1!=0
, then
+ a circular kernel with diameter size1
is used;
+ otherwise the diameter of the kernel is computed from sigma2
.
Information about bilateral filtering
can be found at
http://www.dai.ed.ac.uk/CVonline/LOCAL_COPIES/MANDUCHI1/Bilateral_Filtering.html
param2
is zero, it is set to param1
.
- When not 0, it should be odd too.
-size2
is zero, it is set to size1
.
+ When not 0, it should be odd too.
+In case of Gaussian kernel this parameter may specify Gaussian sigma (standard deviation).
If it is zero, it is calculated from the kernel size:
sigma = (n/2 - 1)*0.3 + 0.8, where n=param1 for horizontal kernel, @@ -840,20 +843,29 @@ void cvSmooth( const CvArr* src, CvArr* dst,With the standard sigma for small kernels (3×3 to 7×7) the performance is better. If
param3
is not zero, while param1
and param2
- are zeros, the kernel size is calculated from the sigma (to provide accurate enough operation).
-param3
) sigma in the vertical direction.
+ are zeros, the kernel size is calculated from the sigma (to provide accurate enough operation).
+
+ In case of Bilateral filter the parameter specifies color sigma; the larger the value, + the stronger the pasterization effect of the filter is.
+In case of non-square Gaussian kernel the parameter may be used to specify a different
+ (from param3
) sigma in the vertical direction.
In case of Bilateral filter the parameter specifies spatial sigma; the larger the value,
+ the stronger the blurring effect of the filter. Note that with large sigma2
+ the processing speed decreases substantionally, so it is recommended to limit the kernel
+ size using the parameter size1
.
-The function cvSmooth
smoothes image using one of several methods. Every of the methods
-has some features and restrictions listed below
Blur with no scaling works with single-channel images only and supports accumulation of
+The function cvSmooth
smoothes image using one of the pre-defined methods. Every of the methods
+has some features and restrictions listed below:
-Simple blur and Gaussian blur support 1- or 3-channel, 8-bit, 16-bit and 32-bit floating point images. -These two methods can process images in-place.
-Median and bilateral filters work with 1- or 3-channel 8-bit images and can not process images -in-place.
- +and accumulation of 32f to 32f format.Applies linear filter to image
@@ -3785,6 +3797,68 @@ the readiness flag for the corresponding image can be set in the next call (i.e. for all the image pairs except the very first oneCV_LKFLOW_PYR_A_READY
is set).
+Constructs a tree of feature vectors
++CvFeatureTree* cvCreateFeatureTree(CvMat* desc); +
+The function cvCreateFeatureTree
constructs a balanced kd-tree index of
+the given feature vectors. The lifetime of the desc matrix must exceed that
+of the returned tree. I.e., no copy is made of the vectors.
+
Destroys a tree of feature vectors
++void cvReleaseFeatureTree(CvFeatureTree* tr); +
+The function cvReleaseFeatureTree
deallocates the given kd-tree.
+
Finds approximate k nearest neighbors of given vectors using best-bin-first search
++void cvFindFeatures(CvFeatureTree* tr, CvMat* desc, + CvMat* results, CvMat* dist, int k=2, int emax=20); +
+The function cvFindFeatures
finds (with high probability) the k
nearest
+neighbors in tr
for each of the given (row-)vectors in desc
, using
+best-bin-first searching ([Beis97]).
+The complexity of the entire operation is at most O(m*emax*log2(n))
,
+where n
is the number of vectors in the tree.
+
Orthogonal range search
++int cvFindFeaturesBoxed(CvFeatureTree* tr, + CvMat* bounds_min, CvMat* bounds_max, + CvMat* results); +
+The function cvFindFeaturesBoxed
performs orthogonal range seaching on the
+given kd-tree. That is, it returns the set of vectors v
in tr
that satisfy
+bounds_min[i] <= v[i] <= bounds_max[i], 0 <= i < d
, where d
is the dimension
+of vectors in the tree.
+The function returns the number of such vectors found.
+
Finds intrinsic and extrinsic camera parameters using calibration pattern
++void cvCalibrationMatrixValues( const CvMat *calibMatr, + int imgWidth, int imgHeight, + double apertureWidth=0, double apertureHeight=0, + double *fovx=NULL, double *fovy=NULL, + double *focalLength=NULL, + CvPoint2D64f *principalPoint=NULL, + double *pixelAspectRatio=NULL ); +
+The function cvCalibrationMatrixValues
computes various useful camera (sensor/lens)
+characteristics using the computed camera calibration matrix, image frame resolution in pixels
+and the physical aperture size.
Finds extrinsic camera parameters for particular view
@@ -5220,72 +5321,75 @@ to use multi-channelNx1
or1xN
arrays.
Nx1
or 1xN
arrays.
Nx1
or 1xN
arrays.