python 與 C++ dlib人臉檢測結果對比,供大家參考,具體內容如下
說明:
由于項目需求發現Linux下c++使用dlib進行人臉檢測和python使用dlib檢測,得到的結果出入比較大,于是寫了測試用例,發現影響結果的原因有但不限于:
1.dlib版本不同(影響不大,幾個像素的差別)
2.dlib 人臉檢測中detector()第二個參數的設置測試結果如下:
python
PDlib.py:
# -*- coding: utf-8 -*-import sysimport cv2 import dlibfrom skimage import iodetector = dlib.get_frontal_face_detector()win = dlib.image_window()for f in sys.argv[1:]: img = io.imread(f) dets = detector(img,1) #使用detector進行人臉檢測 for i, d in enumerate(dets): x = d.left() y = d.top() w = d.right() h = d.bottom() cv2.rectangle(img, (x, y), (w, h), (0, 255, 0)) print("({},{},{},{})".format( x, y, (w-x), (h-y))) win.set_image(img) io.imsave('./P_Dlib_test.jpg',img) #等待點擊 dlib.hit_enter_to_continue()
C++
CDlib.cpp:
#include <dlib/image_processing/frontal_face_detector.h>#include <dlib/opencv.h>#include "opencv2/opencv.hpp"#include <iostream>using namespace dlib;using namespace std;cv::Rect Detect(cv::Mat im){ cv::Rect R; frontal_face_detector detector = get_frontal_face_detector(); array2d<bgr_pixel> img; assign_image(img, cv_image<uchar>(im)); std::vector<rectangle> dets = detector(img);//檢測人臉 //查找最大臉 if (dets.size() != 0) { int Max = 0; int area = 0; for (unsigned long t = 0; t < dets.size(); ++t) { if (area < dets[t].width()*dets[t].height()) { area = dets[t].width()*dets[t].height(); Max = t; } } R.x = dets[Max].left(); R.y = dets[Max].top(); R.width = dets[Max].width(); R.height = dets[Max].height(); cout<<"("<<R.x<<","<<R.y<<","<<R.width<<","<<R.height<<")"<<endl; } return R;}int main(int argc, char** argv){ if (argc != 2) { fprintf(stderr, "請輸入正確參數/n"); return 1; } string path = argv[1]; try { cv::Mat src, dec; src = cv::imread(path); src.copyTo(dec); cv::cvtColor(dec, dec, CV_BGR2GRAY); cv::Rect box; box = Detect(dec); cv::rectangle(src, box, cv::Scalar(0, 0, 255), 1, 1, 0); cv::imshow("frame", src); cv::imwrite("./C_Dlib_test.jpg", src); cv::waitKey(0);//等待建入 } catch (exception& e) { cout << e.what() << endl; }}
項目編譯及運行
python
運行腳本 python PDlib.py G:/DlibTest/data/bush.jpg
C++
創建編譯文件夾 mkdir cbuild 切換到編譯目錄 cd cbuild 生成makefile文件 cmake .. 編譯項目 make 運行可執行文件 ./DlibTest G:/DlibTest/data/bush.jpg新聞熱點
疑難解答