linefit.cpp 16.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
/*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"

43 44 45
namespace cv
{

46 47
static const double eps = 1e-6;

48
static void fitLine2D_wods( const Point2f* points, int count, float *weights, float *line )
49 50 51 52 53 54
{
    double x = 0, y = 0, x2 = 0, y2 = 0, xy = 0, w = 0;
    double dx2, dy2, dxy;
    int i;
    float t;

55
    // Calculating the average of x and y...
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
    if( weights == 0 )
    {
        for( i = 0; i < count; i += 1 )
        {
            x += points[i].x;
            y += points[i].y;
            x2 += points[i].x * points[i].x;
            y2 += points[i].y * points[i].y;
            xy += points[i].x * points[i].y;
        }
        w = (float) count;
    }
    else
    {
        for( i = 0; i < count; i += 1 )
        {
            x += weights[i] * points[i].x;
            y += weights[i] * points[i].y;
            x2 += weights[i] * points[i].x * points[i].x;
            y2 += weights[i] * points[i].y * points[i].y;
            xy += weights[i] * points[i].x * points[i].y;
            w += weights[i];
        }
    }

    x /= w;
    y /= w;
    x2 /= w;
    y2 /= w;
    xy /= w;

    dx2 = x2 - x * x;
    dy2 = y2 - y * y;
    dxy = xy - x * y;

    t = (float) atan2( 2 * dxy, dx2 - dy2 ) / 2;
    line[0] = (float) cos( t );
    line[1] = (float) sin( t );

    line[2] = (float) x;
    line[3] = (float) y;
}

99
static void fitLine3D_wods( const Point3f * points, int count, float *weights, float *line )
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
{
    int i;
    float w0 = 0;
    float x0 = 0, y0 = 0, z0 = 0;
    float x2 = 0, y2 = 0, z2 = 0, xy = 0, yz = 0, xz = 0;
    float dx2, dy2, dz2, dxy, dxz, dyz;
    float *v;
    float n;
    float det[9], evc[9], evl[3];

    memset( evl, 0, 3*sizeof(evl[0]));
    memset( evc, 0, 9*sizeof(evl[0]));

    if( weights )
    {
        for( i = 0; i < count; i++ )
        {
            float x = points[i].x;
            float y = points[i].y;
            float z = points[i].z;
            float w = weights[i];


            x2 += x * x * w;
            xy += x * y * w;
            xz += x * z * w;
            y2 += y * y * w;
            yz += y * z * w;
            z2 += z * z * w;
            x0 += x * w;
            y0 += y * w;
            z0 += z * w;
            w0 += w;
        }
    }
    else
    {
        for( i = 0; i < count; i++ )
        {
            float x = points[i].x;
            float y = points[i].y;
            float z = points[i].z;

            x2 += x * x;
            xy += x * y;
            xz += x * z;
            y2 += y * y;
            yz += y * z;
            z2 += z * z;
            x0 += x;
            y0 += y;
            z0 += z;
        }
        w0 = (float) count;
    }

    x2 /= w0;
    xy /= w0;
    xz /= w0;
    y2 /= w0;
    yz /= w0;
    z2 /= w0;

    x0 /= w0;
    y0 /= w0;
    z0 /= w0;

    dx2 = x2 - x0 * x0;
    dxy = xy - x0 * y0;
    dxz = xz - x0 * z0;
    dy2 = y2 - y0 * y0;
    dyz = yz - y0 * z0;
    dz2 = z2 - z0 * z0;

    det[0] = dz2 + dy2;
    det[1] = -dxy;
    det[2] = -dxz;
    det[3] = det[1];
    det[4] = dx2 + dz2;
    det[5] = -dyz;
    det[6] = det[2];
    det[7] = det[5];
    det[8] = dy2 + dx2;

184 185 186 187 188
    // Searching for a eigenvector of det corresponding to the minimal eigenvalue
    Mat _det( 3, 3, CV_32F, det );
    Mat _evc( 3, 3, CV_32F, evc );
    Mat _evl( 3, 1, CV_32F, evl );
    eigen( _det, _evl, _evc );
189 190 191
    i = evl[0] < evl[1] ? (evl[0] < evl[2] ? 0 : 2) : (evl[1] < evl[2] ? 1 : 2);

    v = &evc[i * 3];
192
    n = (float) std::sqrt( (double)v[0] * v[0] + (double)v[1] * v[1] + (double)v[2] * v[2] );
193 194 195 196 197 198 199 200 201
    n = (float)MAX(n, eps);
    line[0] = v[0] / n;
    line[1] = v[1] / n;
    line[2] = v[2] / n;
    line[3] = x0;
    line[4] = y0;
    line[5] = z0;
}

202
static double calcDist2D( const Point2f* points, int count, float *_line, float *dist )
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222
{
    int j;
    float px = _line[2], py = _line[3];
    float nx = _line[1], ny = -_line[0];
    double sum_dist = 0.;

    for( j = 0; j < count; j++ )
    {
        float x, y;

        x = points[j].x - px;
        y = points[j].y - py;

        dist[j] = (float) fabs( nx * x + ny * y );
        sum_dist += dist[j];
    }

    return sum_dist;
}

223
static double calcDist3D( const Point3f* points, int count, float *_line, float *dist )
224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242
{
    int j;
    float px = _line[3], py = _line[4], pz = _line[5];
    float vx = _line[0], vy = _line[1], vz = _line[2];
    double sum_dist = 0.;

    for( j = 0; j < count; j++ )
    {
        float x, y, z;
        double p1, p2, p3;

        x = points[j].x - px;
        y = points[j].y - py;
        z = points[j].z - pz;

        p1 = vy * z - vz * y;
        p2 = vz * x - vx * z;
        p3 = vx * y - vy * x;

243
        dist[j] = (float) std::sqrt( p1*p1 + p2*p2 + p3*p3 );
244 245 246 247 248 249
        sum_dist += dist[j];
    }

    return sum_dist;
}

250
static void weightL1( float *d, int count, float *w )
251 252 253 254 255 256 257 258 259 260
{
    int i;

    for( i = 0; i < count; i++ )
    {
        double t = fabs( (double) d[i] );
        w[i] = (float)(1. / MAX(t, eps));
    }
}

261
static void weightL12( float *d, int count, float *w )
262 263 264 265 266
{
    int i;

    for( i = 0; i < count; i++ )
    {
267
        w[i] = 1.0f / (float) std::sqrt( 1 + (double) (d[i] * d[i] * 0.5) );
268 269 270 271
    }
}


272
static void weightHuber( float *d, int count, float *w, float _c )
273 274 275 276 277 278 279 280 281 282 283 284 285 286
{
    int i;
    const float c = _c <= 0 ? 1.345f : _c;

    for( i = 0; i < count; i++ )
    {
        if( d[i] < c )
            w[i] = 1.0f;
        else
            w[i] = c/d[i];
    }
}


287
static void weightFair( float *d, int count, float *w, float _c )
288 289 290 291 292 293 294 295 296 297
{
    int i;
    const float c = _c == 0 ? 1 / 1.3998f : 1 / _c;

    for( i = 0; i < count; i++ )
    {
        w[i] = 1 / (1 + d[i] * c);
    }
}

298
static void weightWelsch( float *d, int count, float *w, float _c )
299 300 301 302 303 304
{
    int i;
    const float c = _c == 0 ? 1 / 2.9846f : 1 / _c;

    for( i = 0; i < count; i++ )
    {
305
        w[i] = (float) std::exp( -d[i] * d[i] * c * c );
306 307 308 309 310
    }
}


/* Takes an array of 2D points, type of distance (including user-defined
311 312 313
 distance specified by callbacks, fills the array of four floats with line
 parameters A, B, C, D, where (A, B) is the normalized direction vector,
 (C, D) is the point that belongs to the line. */
314

315 316
static void fitLine2D( const Point2f * points, int count, int dist,
                      float _param, float reps, float aeps, float *line )
317 318 319 320 321 322 323 324 325
{
    double EPS = count*FLT_EPSILON;
    void (*calc_weights) (float *, int, float *) = 0;
    void (*calc_weights_param) (float *, int, float *, float) = 0;
    int i, j, k;
    float _line[6], _lineprev[6];
    float rdelta = reps != 0 ? reps : 1.0f;
    float adelta = aeps != 0 ? aeps : 0.01f;
    double min_err = DBL_MAX, err = 0;
326
    RNG rng((uint64)-1);
327 328 329 330 331 332

    memset( line, 0, 4*sizeof(line[0]) );

    switch (dist)
    {
    case CV_DIST_L2:
333
        return fitLine2D_wods( points, count, 0, line );
334 335

    case CV_DIST_L1:
336
        calc_weights = weightL1;
337 338 339
        break;

    case CV_DIST_L12:
340
        calc_weights = weightL12;
341 342 343
        break;

    case CV_DIST_FAIR:
344
        calc_weights_param = weightFair;
345 346 347
        break;

    case CV_DIST_WELSCH:
348
        calc_weights_param = weightWelsch;
349 350 351
        break;

    case CV_DIST_HUBER:
352
        calc_weights_param = weightHuber;
353 354
        break;

355 356 357
    /*case DIST_USER:
     calc_weights = (void ( * )(float *, int, float *)) _PFP.fp;
     break;*/
358
    default:
359
        CV_Error(CV_StsBadArg, "Unknown distance type");
360 361
    }

362 363
    AutoBuffer<float> wr(count*2);
    float *w = wr, *r = w + count;
364 365 366 367 368 369

    for( k = 0; k < 20; k++ )
    {
        int first = 1;
        for( i = 0; i < count; i++ )
            w[i] = 0.f;
370

371 372
        for( i = 0; i < MIN(count,10); )
        {
373
            j = rng.uniform(0, count);
374 375 376 377 378 379 380
            if( w[j] < FLT_EPSILON )
            {
                w[j] = 1.f;
                i++;
            }
        }

381
        fitLine2D_wods( points, count, w, _line );
382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
        for( i = 0; i < 30; i++ )
        {
            double sum_w = 0;

            if( first )
            {
                first = 0;
            }
            else
            {
                double t = _line[0] * _lineprev[0] + _line[1] * _lineprev[1];
                t = MAX(t,-1.);
                t = MIN(t,1.);
                if( fabs(acos(t)) < adelta )
                {
                    float x, y, d;

                    x = (float) fabs( _line[2] - _lineprev[2] );
                    y = (float) fabs( _line[3] - _lineprev[3] );

                    d = x > y ? x : y;
                    if( d < rdelta )
                        break;
                }
            }
            /* calculate distances */
408
            err = calcDist2D( points, count, _line, r );
409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436
            if( err < EPS )
                break;

            /* calculate weights */
            if( calc_weights )
                calc_weights( r, count, w );
            else
                calc_weights_param( r, count, w, _param );

            for( j = 0; j < count; j++ )
                sum_w += w[j];

            if( fabs(sum_w) > FLT_EPSILON )
            {
                sum_w = 1./sum_w;
                for( j = 0; j < count; j++ )
                    w[j] = (float)(w[j]*sum_w);
            }
            else
            {
                for( j = 0; j < count; j++ )
                    w[j] = 1.f;
            }

            /* save the line parameters */
            memcpy( _lineprev, _line, 4 * sizeof( float ));

            /* Run again... */
437
            fitLine2D_wods( points, count, w, _line );
438 439 440 441 442 443 444 445 446 447 448 449 450 451
        }

        if( err < min_err )
        {
            min_err = err;
            memcpy( line, _line, 4 * sizeof(line[0]));
            if( err < EPS )
                break;
        }
    }
}


/* Takes an array of 3D points, type of distance (including user-defined
452 453 454 455 456
 distance specified by callbacks, fills the array of four floats with line
 parameters A, B, C, D, E, F, where (A, B, C) is the normalized direction vector,
 (D, E, F) is the point that belongs to the line. */
static void fitLine3D( Point3f * points, int count, int dist,
                       float _param, float reps, float aeps, float *line )
457 458 459 460 461
{
    double EPS = count*FLT_EPSILON;
    void (*calc_weights) (float *, int, float *) = 0;
    void (*calc_weights_param) (float *, int, float *, float) = 0;
    int i, j, k;
462
    float _line[6]={0,0,0,0,0,0}, _lineprev[6]={0,0,0,0,0,0};
463 464 465
    float rdelta = reps != 0 ? reps : 1.0f;
    float adelta = aeps != 0 ? aeps : 0.01f;
    double min_err = DBL_MAX, err = 0;
466
    RNG rng((uint64)-1);
467 468 469 470

    switch (dist)
    {
    case CV_DIST_L2:
471
        return fitLine3D_wods( points, count, 0, line );
472 473

    case CV_DIST_L1:
474
        calc_weights = weightL1;
475 476 477
        break;

    case CV_DIST_L12:
478
        calc_weights = weightL12;
479 480 481
        break;

    case CV_DIST_FAIR:
482
        calc_weights_param = weightFair;
483 484 485
        break;

    case CV_DIST_WELSCH:
486
        calc_weights_param = weightWelsch;
487 488 489
        break;

    case CV_DIST_HUBER:
490
        calc_weights_param = weightHuber;
491 492 493
        break;

    default:
494
        CV_Error(CV_StsBadArg, "Unknown distance");
495 496
    }

497 498
    AutoBuffer<float> buf(count*2);
    float *w = buf, *r = w + count;
499 500 501 502 503 504

    for( k = 0; k < 20; k++ )
    {
        int first = 1;
        for( i = 0; i < count; i++ )
            w[i] = 0.f;
505

506 507
        for( i = 0; i < MIN(count,10); )
        {
508
            j = rng.uniform(0, count);
509 510 511 512 513 514 515
            if( w[j] < FLT_EPSILON )
            {
                w[j] = 1.f;
                i++;
            }
        }

516
        fitLine3D_wods( points, count, w, _line );
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549
        for( i = 0; i < 30; i++ )
        {
            double sum_w = 0;

            if( first )
            {
                first = 0;
            }
            else
            {
                double t = _line[0] * _lineprev[0] + _line[1] * _lineprev[1] + _line[2] * _lineprev[2];
                t = MAX(t,-1.);
                t = MIN(t,1.);
                if( fabs(acos(t)) < adelta )
                {
                    float x, y, z, ax, ay, az, dx, dy, dz, d;

                    x = _line[3] - _lineprev[3];
                    y = _line[4] - _lineprev[4];
                    z = _line[5] - _lineprev[5];
                    ax = _line[0] - _lineprev[0];
                    ay = _line[1] - _lineprev[1];
                    az = _line[2] - _lineprev[2];
                    dx = (float) fabs( y * az - z * ay );
                    dy = (float) fabs( z * ax - x * az );
                    dz = (float) fabs( x * ay - y * ax );

                    d = dx > dy ? (dx > dz ? dx : dz) : (dy > dz ? dy : dz);
                    if( d < rdelta )
                        break;
                }
            }
            /* calculate distances */
550 551 552
            err = calcDist3D( points, count, _line, r );
            //if( err < FLT_EPSILON*count )
            //    break;
553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573

            /* calculate weights */
            if( calc_weights )
                calc_weights( r, count, w );
            else
                calc_weights_param( r, count, w, _param );

            for( j = 0; j < count; j++ )
                sum_w += w[j];

            if( fabs(sum_w) > FLT_EPSILON )
            {
                sum_w = 1./sum_w;
                for( j = 0; j < count; j++ )
                    w[j] = (float)(w[j]*sum_w);
            }
            else
            {
                for( j = 0; j < count; j++ )
                    w[j] = 1.f;
            }
574

575 576
            /* save the line parameters */
            memcpy( _lineprev, _line, 6 * sizeof( float ));
577

578
            /* Run again... */
579
            fitLine3D_wods( points, count, w, _line );
580
        }
581

582 583 584 585 586 587 588 589 590 591
        if( err < min_err )
        {
            min_err = err;
            memcpy( line, _line, 6 * sizeof(line[0]));
            if( err < EPS )
                break;
        }
    }
}

592
}
593

594 595
void cv::fitLine( InputArray _points, OutputArray _line, int distType,
                 double param, double reps, double aeps )
596
{
597 598
    CV_INSTRUMENT_REGION()

599
    Mat points = _points.getMat();
600

601 602 603
    float linebuf[6]={0.f};
    int npoints2 = points.checkVector(2, -1, false);
    int npoints3 = points.checkVector(3, -1, false);
604

605
    CV_Assert( npoints2 >= 0 || npoints3 >= 0 );
606

607
    if( points.depth() != CV_32F || !points.isContinuous() )
608
    {
609 610 611
        Mat temp;
        points.convertTo(temp, CV_32F);
        points = temp;
612
    }
613

614 615 616
    if( npoints2 >= 0 )
        fitLine2D( points.ptr<Point2f>(), npoints2, distType,
                   (float)param, (float)reps, (float)aeps, linebuf);
617
    else
618 619
        fitLine3D( points.ptr<Point3f>(), npoints3, distType,
                   (float)param, (float)reps, (float)aeps, linebuf);
620

621 622
    Mat(npoints2 >= 0 ? 4 : 6, 1, CV_32F, linebuf).copyTo(_line);
}
623 624


625 626 627 628 629
CV_IMPL void
cvFitLine( const CvArr* array, int dist, double param,
           double reps, double aeps, float *line )
{
    CV_Assert(line != 0);
630

631 632 633
    cv::AutoBuffer<double> buf;
    cv::Mat points = cv::cvarrToMat(array, false, false, 0, &buf);
    cv::Mat linemat(points.checkVector(2) >= 0 ? 4 : 6, 1, CV_32F, line);
634

635
    cv::fitLine(points, linemat, dist, param, reps, aeps);
636 637 638
}

/* End of file. */