mirror of https://github.com/opencv/opencv.git
Open Source Computer Vision Library
https://opencv.org/
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
402 lines
12 KiB
402 lines
12 KiB
/*M/////////////////////////////////////////////////////////////////////////////////////// |
|
// |
|
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING. |
|
// |
|
// By downloading, copying, installing or using the software you agree to this license. |
|
// If you do not agree to this license, do not download, install, |
|
// copy or use the software. |
|
// |
|
// |
|
// Intel License Agreement |
|
// For Open Source Computer Vision Library |
|
// |
|
// Copyright (C) 2000, Intel Corporation, all rights reserved. |
|
// Third party copyrights are property of their respective owners. |
|
// |
|
// Redistribution and use in source and binary forms, with or without modification, |
|
// are permitted provided that the following conditions are met: |
|
// |
|
// * Redistribution's of source code must retain the above copyright notice, |
|
// this list of conditions and the following disclaimer. |
|
// |
|
// * Redistribution's in binary form must reproduce the above copyright notice, |
|
// this list of conditions and the following disclaimer in the documentation |
|
// and/or other materials provided with the distribution. |
|
// |
|
// * The name of Intel Corporation may not be used to endorse or promote products |
|
// derived from this software without specific prior written permission. |
|
// |
|
// This software is provided by the copyright holders and contributors "as is" and |
|
// any express or implied warranties, including, but not limited to, the implied |
|
// warranties of merchantability and fitness for a particular purpose are disclaimed. |
|
// In no event shall the Intel Corporation or contributors be liable for any direct, |
|
// indirect, incidental, special, exemplary, or consequential damages |
|
// (including, but not limited to, procurement of substitute goods or services; |
|
// loss of use, data, or profits; or business interruption) however caused |
|
// and on any theory of liability, whether in contract, strict liability, |
|
// or tort (including negligence or otherwise) arising in any way out of |
|
// the use of this software, even if advised of the possibility of such damage. |
|
// |
|
//M*/ |
|
#include "precomp.hpp" |
|
|
|
typedef struct _PointInfo |
|
{ |
|
CvPoint pt; |
|
int left_neigh; |
|
int right_neigh; |
|
|
|
} |
|
icvPointInfo; |
|
|
|
|
|
static CvStatus |
|
icvFindDominantPointsIPAN( CvSeq * contour, |
|
CvMemStorage * storage, |
|
CvSeq ** corners, int dmin2, int dmax2, int dneigh2, float amax ) |
|
{ |
|
CvStatus status = CV_OK; |
|
|
|
/* variables */ |
|
int n = contour->total; |
|
|
|
float *sharpness; |
|
float *distance; |
|
icvPointInfo *ptInf; |
|
|
|
int i, j, k; |
|
|
|
CvSeqWriter writer; |
|
|
|
float mincos = (float) cos( 3.14159265359 * amax / 180 ); |
|
|
|
/* check bad arguments */ |
|
if( contour == NULL ) |
|
return CV_NULLPTR_ERR; |
|
if( storage == NULL ) |
|
return CV_NULLPTR_ERR; |
|
if( corners == NULL ) |
|
return CV_NULLPTR_ERR; |
|
if( dmin2 < 0 ) |
|
return CV_BADSIZE_ERR; |
|
if( dmax2 < dmin2 ) |
|
return CV_BADSIZE_ERR; |
|
if( (dneigh2 > dmax2) || (dneigh2 < 0) ) |
|
return CV_BADSIZE_ERR; |
|
if( (amax < 0) || (amax > 180) ) |
|
return CV_BADSIZE_ERR; |
|
|
|
sharpness = (float *) cvAlloc( n * sizeof( float )); |
|
distance = (float *) cvAlloc( n * sizeof( float )); |
|
|
|
ptInf = (icvPointInfo *) cvAlloc( n * sizeof( icvPointInfo )); |
|
|
|
/*****************************************************************************************/ |
|
/* First pass */ |
|
/*****************************************************************************************/ |
|
|
|
if( CV_IS_SEQ_CHAIN_CONTOUR( contour )) |
|
{ |
|
CvChainPtReader reader; |
|
|
|
cvStartReadChainPoints( (CvChain *) contour, &reader ); |
|
|
|
for( i = 0; i < n; i++ ) |
|
{ |
|
CV_READ_CHAIN_POINT( ptInf[i].pt, reader ); |
|
} |
|
} |
|
else if( CV_IS_SEQ_POINT_SET( contour )) |
|
{ |
|
CvSeqReader reader; |
|
|
|
cvStartReadSeq( contour, &reader, 0 ); |
|
|
|
for( i = 0; i < n; i++ ) |
|
{ |
|
CV_READ_SEQ_ELEM( ptInf[i].pt, reader ); |
|
} |
|
} |
|
else |
|
{ |
|
return CV_BADFLAG_ERR; |
|
} |
|
|
|
for( i = 0; i < n; i++ ) |
|
{ |
|
/* find nearest suitable points |
|
which satisfy distance constraint >dmin */ |
|
int left_near = 0; |
|
int right_near = 0; |
|
int left_far, right_far; |
|
|
|
float dist_l = 0; |
|
float dist_r = 0; |
|
|
|
int i_plus = 0; |
|
int i_minus = 0; |
|
|
|
float max_cos_alpha; |
|
|
|
/* find right minimum */ |
|
while( dist_r < dmin2 ) |
|
{ |
|
float dx, dy; |
|
int ind; |
|
|
|
if( i_plus >= n ) |
|
goto error; |
|
|
|
right_near = i_plus; |
|
|
|
if( dist_r < dneigh2 ) |
|
ptInf[i].right_neigh = i_plus; |
|
|
|
i_plus++; |
|
|
|
ind = (i + i_plus) % n; |
|
dx = (float) (ptInf[i].pt.x - ptInf[ind].pt.x); |
|
dy = (float) (ptInf[i].pt.y - ptInf[ind].pt.y); |
|
dist_r = dx * dx + dy * dy; |
|
} |
|
/* find right maximum */ |
|
while( dist_r <= dmax2 ) |
|
{ |
|
float dx, dy; |
|
int ind; |
|
|
|
if( i_plus >= n ) |
|
goto error; |
|
|
|
distance[(i + i_plus) % n] = cvSqrt( dist_r ); |
|
|
|
if( dist_r < dneigh2 ) |
|
ptInf[i].right_neigh = i_plus; |
|
|
|
i_plus++; |
|
|
|
right_far = i_plus; |
|
|
|
ind = (i + i_plus) % n; |
|
|
|
dx = (float) (ptInf[i].pt.x - ptInf[ind].pt.x); |
|
dy = (float) (ptInf[i].pt.y - ptInf[ind].pt.y); |
|
dist_r = dx * dx + dy * dy; |
|
} |
|
right_far = i_plus; |
|
|
|
/* left minimum */ |
|
while( dist_l < dmin2 ) |
|
{ |
|
float dx, dy; |
|
int ind; |
|
|
|
if( i_minus <= -n ) |
|
goto error; |
|
|
|
left_near = i_minus; |
|
|
|
if( dist_l < dneigh2 ) |
|
ptInf[i].left_neigh = i_minus; |
|
|
|
i_minus--; |
|
|
|
ind = i + i_minus; |
|
ind = (ind < 0) ? (n + ind) : ind; |
|
|
|
dx = (float) (ptInf[i].pt.x - ptInf[ind].pt.x); |
|
dy = (float) (ptInf[i].pt.y - ptInf[ind].pt.y); |
|
dist_l = dx * dx + dy * dy; |
|
} |
|
|
|
/* find left maximum */ |
|
while( dist_l <= dmax2 ) |
|
{ |
|
float dx, dy; |
|
int ind; |
|
|
|
if( i_minus <= -n ) |
|
goto error; |
|
|
|
ind = i + i_minus; |
|
ind = (ind < 0) ? (n + ind) : ind; |
|
|
|
distance[ind] = cvSqrt( dist_l ); |
|
|
|
if( dist_l < dneigh2 ) |
|
ptInf[i].left_neigh = i_minus; |
|
|
|
i_minus--; |
|
|
|
left_far = i_minus; |
|
|
|
ind = i + i_minus; |
|
ind = (ind < 0) ? (n + ind) : ind; |
|
|
|
dx = (float) (ptInf[i].pt.x - ptInf[ind].pt.x); |
|
dy = (float) (ptInf[i].pt.y - ptInf[ind].pt.y); |
|
dist_l = dx * dx + dy * dy; |
|
} |
|
left_far = i_minus; |
|
|
|
if( (i_plus - i_minus) > n + 2 ) |
|
goto error; |
|
|
|
max_cos_alpha = -1; |
|
for( j = left_far + 1; j < left_near; j++ ) |
|
{ |
|
float dx, dy; |
|
float a, a2; |
|
int leftind = i + j; |
|
|
|
leftind = (leftind < 0) ? (n + leftind) : leftind; |
|
|
|
a = distance[leftind]; |
|
a2 = a * a; |
|
|
|
for( k = right_near + 1; k < right_far; k++ ) |
|
{ |
|
int ind = (i + k) % n; |
|
float c2, cosalpha; |
|
float b = distance[ind]; |
|
float b2 = b * b; |
|
|
|
/* compute cosinus */ |
|
dx = (float) (ptInf[leftind].pt.x - ptInf[ind].pt.x); |
|
dy = (float) (ptInf[leftind].pt.y - ptInf[ind].pt.y); |
|
|
|
c2 = dx * dx + dy * dy; |
|
cosalpha = (a2 + b2 - c2) / (2 * a * b); |
|
|
|
max_cos_alpha = MAX( max_cos_alpha, cosalpha ); |
|
|
|
if( max_cos_alpha < mincos ) |
|
max_cos_alpha = -1; |
|
|
|
sharpness[i] = max_cos_alpha; |
|
} |
|
} |
|
} |
|
/*****************************************************************************************/ |
|
/* Second pass */ |
|
/*****************************************************************************************/ |
|
|
|
cvStartWriteSeq( (contour->flags & ~CV_SEQ_ELTYPE_MASK) | CV_SEQ_ELTYPE_INDEX, |
|
sizeof( CvSeq ), sizeof( int ), storage, &writer ); |
|
|
|
/* second pass - nonmaxima suppression */ |
|
/* neighborhood of point < dneigh2 */ |
|
for( i = 0; i < n; i++ ) |
|
{ |
|
int suppressed = 0; |
|
if( sharpness[i] == -1 ) |
|
continue; |
|
|
|
for( j = 1; (j <= ptInf[i].right_neigh) && (suppressed == 0); j++ ) |
|
{ |
|
if( sharpness[i] < sharpness[(i + j) % n] ) |
|
suppressed = 1; |
|
} |
|
|
|
for( j = -1; (j >= ptInf[i].left_neigh) && (suppressed == 0); j-- ) |
|
{ |
|
int ind = i + j; |
|
|
|
ind = (ind < 0) ? (n + ind) : ind; |
|
if( sharpness[i] < sharpness[ind] ) |
|
suppressed = 1; |
|
} |
|
|
|
if( !suppressed ) |
|
CV_WRITE_SEQ_ELEM( i, writer ); |
|
} |
|
|
|
*corners = cvEndWriteSeq( &writer ); |
|
|
|
cvFree( &sharpness ); |
|
cvFree( &distance ); |
|
cvFree( &ptInf ); |
|
|
|
return status; |
|
|
|
error: |
|
/* dmax is so big (more than contour diameter) |
|
that algorithm could become infinite cycle */ |
|
cvFree( &sharpness ); |
|
cvFree( &distance ); |
|
cvFree( &ptInf ); |
|
|
|
return CV_BADRANGE_ERR; |
|
} |
|
|
|
|
|
/*F/////////////////////////////////////////////////////////////////////////////////////// |
|
// Name: icvFindDominantPoints |
|
// Purpose: |
|
// Applies some algorithm to find dominant points ( corners ) of contour |
|
// |
|
// Context: |
|
// Parameters: |
|
// contours - pointer to input contour object. |
|
// out_numbers - array of dominant points indices |
|
// count - length of out_numbers array on input |
|
// and numbers of founded dominant points on output |
|
// |
|
// method - only CV_DOMINANT_IPAN now |
|
// parameters - array of parameters |
|
// for IPAN algorithm |
|
// [0] - minimal distance |
|
// [1] - maximal distance |
|
// [2] - neighborhood distance (must be not greater than dmaximal distance) |
|
// [3] - maximal possible angle of curvature |
|
// Returns: |
|
// CV_OK or error code |
|
// Notes: |
|
// User must allocate out_numbers array. If it is small - function fills array |
|
// with part of points and returns error |
|
//F*/ |
|
CV_IMPL CvSeq* |
|
cvFindDominantPoints( CvSeq * contour, CvMemStorage * storage, int method, |
|
double parameter1, double parameter2, double parameter3, double parameter4 ) |
|
{ |
|
CvSeq* corners = 0; |
|
|
|
if( !contour ) |
|
CV_Error( CV_StsNullPtr, "" ); |
|
|
|
if( !storage ) |
|
storage = contour->storage; |
|
|
|
if( !storage ) |
|
CV_Error( CV_StsNullPtr, "" ); |
|
|
|
switch (method) |
|
{ |
|
case CV_DOMINANT_IPAN: |
|
{ |
|
int dmin = cvRound(parameter1); |
|
int dmax = cvRound(parameter2); |
|
int dneigh = cvRound(parameter3); |
|
int amax = cvRound(parameter4); |
|
|
|
if( amax == 0 ) |
|
amax = 150; |
|
if( dmin == 0 ) |
|
dmin = 7; |
|
if( dmax == 0 ) |
|
dmax = dmin + 2; |
|
if( dneigh == 0 ) |
|
dneigh = dmin; |
|
|
|
IPPI_CALL( icvFindDominantPointsIPAN( contour, storage, &corners, |
|
dmin*dmin, dmax*dmax, dneigh*dneigh, (float)amax )); |
|
} |
|
break; |
|
default: |
|
CV_Error( CV_StsBadArg, "" ); |
|
} |
|
|
|
return corners; |
|
} |
|
|
|
/* End of file. */
|
|
|