扫描线－－通用多边形填充算法

扫描线－－通用多边形填充算法
该算法有几个可学习的地方：
（1）正负1思想
（2）对边界条件的处理
（3）数据结构的选择

code:
sweep.h

#ifndef SWEEP_H
#define SWEEP_H

struct Edge {
    int nxty;
    int curx;
    int dx, dy; // 所在扫描线的增量
    Edge nxt;
};

//扫描线主算法
void sweep(int p[][2], int n, void (setPixel)(int, int));
#endifsweep.cpp
#include “sweep.h”
#include <algorithm>
using namespace std;

const int MAXN = 1024;

int cp[MAXN][2], n;
inline bool cmp(int i, int j) {
    return cp[i][1] < cp[j][1]

(cp[i][1] == cp[j][1] && cp[i][0] < cp[j][0]);
}

Edge * e[MAXN], *h, *ph, *data;
void insert(int ly, int px, int ind) {
    int y1,y2,y, nxt, pre, flag=0;

    nxt = (ind + 1) % n; pre = (ind - 1 + n) % n;
    y = cp[ind][1]; y1 = cp[nxt][1]; y2 = cp[pre][1];
    if (y1 > y2) swap(y1, y2);
    if (y1 < y && y < y2) {
        //需缩短一个单位
        flag = 1;
    }

    h = e[ly]; ph=NULL;
    while (h) {
        if (h->dy > cp[ind][1]

(h->dy == cp[ind][1] && h->dx > cp[ind][0])) break;
        ph = h;
        h = h->nxt;
    }

    data = new Edge;
    data->curx = px; data->nxty = cp[ind][1]; data->dx = cp[ind][0] - px; data->dy = cp[ind][1] - ly; data->nxt = NULL;
    if (flag) data->nxty–;

    if (ph) {
        data->nxt = ph->nxt;
        ph->nxt = data;
    } else {
        data->nxt = e[ly];
        e[ly] = data;
    }


}

int ex[MAXN][MAXN], ne[MAXN];
inline int abs(int a) {
    return a > 0 ? a : -a;
}
void makepoint(int line, Edge h) {
    int dx = h->dx, dy = h->dy, cnt=0;
    int x, y, flag=1;
    if ((h->dx)(h->dy)<0) flag=0;
    for (y=line, x=h->curx; y<=h->nxty; y++) {
        ex[y][ne[y]++] = x;
        cnt += 2abs(dx);
        while (cnt>=2abs(dy)) {
            cnt -= 2abs(dy);
            if (flag) x++;
            else x–;
        }
    }
}

void sweep(int p[][2], int nn, void (setPixel)(int, int)) {
    //对所有点按y坐标递增排序,y坐标相等的按x坐标递增排序
    n = nn;
    int i, j, k, ind, nxt, pre;
    int *num = new int[n];    //点索引;
    for (i=0; i<n; i++) num[i] = i;
    memcpy(cp, p, sizeof(cp));
    sort(num, num+n, cmp);

    //建立有序边表
    memset(e, 0, sizeof(e));
    for (i=0; i<n; i++) {
        ind = num[i];
        nxt = (ind + 1) % n;
        pre = (ind - 1 + n) % n;
        if (p[nxt][1] > p[ind][1]) insert(p[ind][1], p[ind][0], nxt);
        if (p[pre][1] > p[ind][1]) insert(p[ind][1], p[ind][0], pre);
    }

    //处理active edge list
    memset(ne, 0, sizeof(ne));
    for (i=0; i<MAXN; i++) {
        h = e[i]; ph = NULL;

        while (h) {
            makepoint(i, h);
            h = h->nxt;
        }
        sort(ex[i], ex[i]+ne[i]);
        for (j=0; j<ne[i]; j+=2)
            for (k=ex[i][j]; k<=ex[i][j+1]; k++)
                setPixel(k,i);
    }

}sweepline.cpp
#include <stdlib.h>
#include <stdio.h>
#include <GL/glut.h>
#include “sweep.h”

void myInit();

void setPixel(int x, int y);

void myDisplay();

int main(int argc, char **argv) {
    glutInit(&argc, argv);
    glutInitDisplayMode(GLUT_SINGLE

GLUT_RGB);
    glutInitWindowSize(640, 480);
    glutInitWindowPosition (100, 150);
    glutCreateWindow(“SweepLine”);
    glutDisplayFunc(myDisplay);
    myInit();
    glutMainLoop();
    return 0;
}

void setPixel(int x, int y) {
    glBegin(GL_POINTS);
    glVertex2i(x, y);
    glEnd();
}

void myInit() {
    glClearColor(1.0, 1.0, 1.0, 0.0);
    glColor3f(0.0, 0.0, 0.0);
    glMatrixMode(GL_PROJECTION);
    glLoadIdentity();
    gluOrtho2D(0.0, 640.0, 0.0, 480.0);
}

void myDisplay() {
    int i, j;
    glClear(GL_COLOR_BUFFER_BIT);
    int p[5][2];
    p[0][0] = 100; p[0][1] = 300;
    p[1][0] = 200; p[1][1] = 50;
    p[2][0] = 300; p[2][1] = 100;
    p[3][0] = 400; p[3][1] = 0;
    p[4][0] = 350; p[4][1] = 470;
    sweep(p, 5, setPixel);
    glFlush();
}

http://zhidao.baidu.com/question/40582214.html

Published on Nov 16, 2009 in categories vision

上一篇 Why most popular javascript libraries doesn’t support xml with namespaces
下一篇 Linux安装与启动