Article Directory
1. Code
Sampling + search, the search here is dijkstra, and Astar can also be used
#include <iostream>
#include <vector>
#include <string>
#include <queue>
#include "opencv2/highgui.hpp"
#include "opencv2/opencv.hpp"
// #include <direct.h>
#include <unistd.h> //目录
#include <typeinfo>
#define INF 13421772
#define sampleNum 200
#define START 0
#define GOAL 1
double toDegree(double radian) {
return radian * 180 / M_PI;}
double toRadian(double degree) {
return degree * M_PI / 180;}
struct GraphNode{
int label;
std::vector<GraphNode*> neighbors;
GraphNode(int x): label(x) {
};
};
void loadMap(){
// char *root_path = NULL;
// root_path = getcwd(NULL,0);
// std::cout << root_path << std::endl;
// cv::Mat dismap = cv::imread("../map/map_2.bmp");
// cv::imshow("dismap", dismap);
// cv::waitKey(1000);
}
bool checkCollision(const std::vector<int> point, const cv::Mat map){
bool reach = true;
if(map.at<cv::Vec3b>(point[1],point[0])[0]==0 &&
map.at<cv::Vec3b>(point[1],point[0])[1]==0 &&
map.at<cv::Vec3b>(point[1],point[0])[2]==0){
reach = false;
}
return reach;
}
//这里只是设置了10个点,也可以根据间距最大的设置采样点
bool checkPath(const std::vector<int> point_a,
const std::vector<int> point_b,
const cv::Mat map, int split_num){
std::vector<double> path_x;
std::vector<double> path_y;
// path_x.reserve(split_num + 1);
// path_y.reserve(split_num + 1);
// int min_x = point_a[0] > point_b[0] ? point_a[0] : point_b[0];
// int min_y = point_a[1] > point_b[1] ? point_a[1] : point_b[1];
double interval_x = (point_b[0] - point_a[0]) / (split_num);
double interval_y = (point_b[1] - point_a[1]) / (split_num);
// std::cout << point_a[0] << "," << point_b[0] << std::endl;
for(int i = 0; i <= split_num; ++i){
path_x.push_back(point_a[0] + i * interval_x);
path_y.push_back(point_a[1] + i * interval_y);
// std::cout << point_a[0] + i * interval_x << "," << point_a[1] + i * interval_y << std::endl;
// std::cout << "i:" << i << std::endl;
}
for(int i = 0; i <= split_num; ++i){
if(!checkCollision({
int(path_x[i]), int(path_y[i])}, map)){
return false;
}
}
return true;
}
double calcDistance(const std::vector<int> point_a, const std::vector<int> point_b){
return sqrt(pow(point_a[0]-point_b[0],2) + pow(point_a[1]-point_b[1],2));
}
struct Dijkstra {
struct node{
int point;
double value;
node(int _point, double _value):point(_point), value(_value){
}
bool operator < (const node& rhs) const{
return value > rhs.value;
}
};
std::vector<node> edges[sampleNum];
double dist[sampleNum];
int path[sampleNum];
void init()
{
for(int i = 0; i < sampleNum; i++)
{
edges[i].clear();
dist[i] = 0;
path[i] = 0;
}
}
void addEdge(int from, int to, double dist)
{
edges[from].push_back(node(to,dist));
// edges[to].push_back(node(from,dist));
}
void showEdge()
{
std::cout << "------------------------" << std::endl;
for(int i = 0; i< sampleNum; i++)
{
// cout << "-----" << i << "-----" << endl;
for(int j=0; j<edges[i].size(); j++)
{
std::cout << i << "," << edges[i][j].point << "," << edges[i][j].value << std::endl;
}
// cout << "**********************" << endl;
}
std::cout << "------------------------" << std::endl;
}
std::vector<int> dijkstra(int s, int t)
{
std::priority_queue <node> q;
for(int i = 0; i < sampleNum; i++)
dist[i] = INF;
dist[s] = 0;
q.push(node(s, dist[s]));
while(!q.empty())
{
node x = q.top(); q.pop();
for (int i = 0; i < edges[x.point].size(); i++)
{
// std::cout << edges[x.point].size() << std::endl;
node y = edges[x.point][i];
if (dist[y.point] > dist[x.point] + y.value)
{
dist[y.point] = dist[x.point] + y.value;
path[y.point] = x.point;
q.push(node(y.point, dist[y.point]));
}
}
}
std::vector<int> result;
// 存距离
// result.push_back(dist[t]);
std::cout << "dist[t]:" << dist[t] << std::endl;
while(t)
{
result.push_back(t);
t = path[t];
}
result.push_back(path[0]);
reverse(result.begin(),result.end()); //起点->终点 ，+ 距离
// for show result (start -> end,cost)
// for(int i=0; i < result.size(); i++)
// {
// std::cout << result[i] << " ";
// }
return result;
}
};
struct Dijkstra DijkstraPlanning;
int main() {
cv::Mat dismap = cv::imread("../map/map_2.bmp");
// cv::imshow("dismap", dismap);
// cv::waitKey(1000);
int mapLength = dismap.cols;
int mapWidth = dismap.rows;
// std::cout << "mapLength:" << mapLength << ",mapWidth:" << mapWidth << std::endl;
// std::cout << dismap.size[0] << typeid(dismap.size).name() << std::endl;
//改为结构体
// std::vector<int> pStrat = {50, 50};
// std::vector<int> pGoal = {450, 450};
std::vector<int> pStrat = {
10, 10};
std::vector<int> pGoal = {
490, 490};
std::vector<std::vector<int> > sampleMap;
sampleMap.reserve(sampleNum + 2);
sampleMap.push_back(pStrat);
sampleMap.push_back(pGoal);
// std::cout << sampleMap.size() << std::endl;
//初始化采样点
while (sampleMap.size() < sampleNum + 2)
// while (sampleMap.size() < 10)
{
std::random_device rd;
std::mt19937 gen(rd());
std::uniform_int_distribution<> dis(0, dismap.size[0]-1);
std::vector<int> res;
// sampleMap.push_back(std::vector<int>());
for(int i=0; i<2; ++i){
res.push_back(dis(gen));
// sampleMap.back().push_back(dis(gen));
}
if(checkCollision(res, dismap)){
sampleMap.push_back({
res[0], res[1]});
}
// std::cout << res[0] << "," << res[1] << std::endl;
}
const int MAX_N = sampleNum + 2;
GraphNode *Graph[MAX_N];
const double DISTANCE = 100;
const int SPLIT_N = 10;
// for(int i = 0; i < MAX_N; i++){
// Graph[i] = new GraphNode(i);
// }
DijkstraPlanning.init();
for (int i = 0; i < MAX_N; i++)
{
for(int j = 0; j < MAX_N; j++){
if (calcDistance(sampleMap[i], sampleMap[j]) <= DISTANCE &&
checkPath(sampleMap[i], sampleMap[j], dismap, SPLIT_N) && i!= j)
{
cv::line(dismap, cv::Point(sampleMap[i][0], sampleMap[i][1]),
cv::Point(sampleMap[j][0], sampleMap[j][1]),
cv::Scalar(0, 0, 255), 1);
// Graph[i]->neighbors.push_back(Graph(i));
DijkstraPlanning.addEdge(i, j, calcDistance(sampleMap[i], sampleMap[j]));
}
}
}
// for(int i=0; i<sampleMap.size(); i++){
// // std::cout << i << ":";
// for(int j=0; j<sampleMap[i].size(); j++){
// // std::cout << sampleMap[i][j] << ",";
// std::cout << sampleMap[i][j] << ",";
// }
// std::cout << std::endl;
// }
DijkstraPlanning.showEdge();
std::vector<int> result = DijkstraPlanning.dijkstra(START, GOAL);
// std::cout << "result.size(): " << result.size() << std::endl;
// std::cout << "result: ";
// for(int i=0; i < result.size(); i++)
// {
// std::cout << result[i] << " ";
// }
if(result.size() > 2){
for(int i=0;i < result.size() - 1; i++)
{
// std::cout << result[i] << " ";
cv::line(dismap, cv::Point(sampleMap[result[i]][0], sampleMap[result[i]][1]),
cv::Point(sampleMap[result[i+1]][0], sampleMap[result[i+1]][1]),
cv::Scalar(0, 255, 0), 5);
}
}
for(int i=0; i<sampleMap.size(); i++){
// std::cout << i << ":";
for(int j=0; j<sampleMap[i].size(); j++){
// std::cout << sampleMap[i][j] << ",";
dismap.at<cv::Vec3b>(sampleMap[i][1],sampleMap[i][0])[0] = 0;
dismap.at<cv::Vec3b>(sampleMap[i][1],sampleMap[i][0])[1] = 255;
dismap.at<cv::Vec3b>(sampleMap[i][1],sampleMap[i][0])[2] = 0;
}
// std::cout << std::endl;
}
// std::cout << sampleMap.size() << std::endl;
cv::imshow("dismap", dismap);
cv::waitKey(10000);
// if(result.size() > 2){
// cv::imwrite("/home/seivl/me/代码/规划/prm/planning_result/prm.bmp", dismap);
// }
return 0;
}
2. Results
