OpenCV实现去除背景识别的方法总结
目录
- 实现效果
- 实现代码
- 补充
实现效果
效果如图,只识别一定距离内的物体
哈哈哈哈哈哈哈哈哈,但我不知道这有什么用
实现代码
import pyrealsense2 as rs import numpy as np import cv2 # 排除背景色 WIDTH = 848 HEIGHT = 480 # 初始化 config = rs.config() config.enable_stream(rs.stream.color, WIDTH, HEIGHT, rs.format.bgr8, 30) config.enable_stream(rs.stream.depth, WIDTH, HEIGHT, rs.format.z16, 30) # 开始 pipeline = rs.pipeline() profile = pipeline.start(config) # 距离[m] = depth * depth_scale depth_sensor = profile.get_device().first_depth_sensor() depth_scale = depth_sensor.get_depth_scale() clipping_distance_in_meters = 0.4 # 40cm以内 clipping_distance = clipping_distance_in_meters / depth_scale # 对齐图像 align_to = rs.stream.color align = rs.align(align_to) threshold = (WIDTH * HEIGHT * 3) * 0.95 try: while True: frames = pipeline.wait_for_frames() aligned_frames = align.process(frames) color_frame = aligned_frames.get_color_frame() depth_frame = aligned_frames.get_depth_frame() if not depth_frame or not color_frame: continue color_image = np.asanyarray(color_frame.get_data()) depth_image = np.asanyarray(depth_frame.get_data()) # clipping_distance_in_metersm以以内形成画像 white_color = 255 # 背景色 depth_image_3d = np.dstack((depth_image, depth_image, depth_image)) bg_removed = np.where((depth_image_3d > clipping_distance) | (depth_image_3d <= 0), white_color, color_image) # 计算具有背景颜色的像素数 white_pic = np.sum(bg_removed == 255) # 当背景颜色低于某个值时显示“检测到” if(threshold > white_pic): print("检测到 {}".format(white_pic)) else: print("{}".format(white_pic)) images = np.hstack((bg_removed, color_image)) cv2.imshow('Frames', images) if cv2.waitKey(1) & 0xff == 27: break finally: # 停止 pipeline.stop() cv2.destroyAllWindows()
补充
在opencv中有两种方法可以进行背景消除:
其一、基于机器学习(Knn–K个最近邻)背景消除建模
其二、于图像分割(GMM,抗干扰图像分割)背景消除建模BS ,Background Subtraction
c版
#include<opencv2/opencv.hpp> #include<iostream> using namespace std; using namespace cv; int main(int argc, char** argv) { VideoCapture capture; capture.open("D:/software/opencv1/picture/vtest.avi"); if (!capture.isOpened()) { printf("could not load the video!"); return -1; } Mat frame; Mat bsmaskMOG2,bsmaskKNN; namedWindow("input video", CV_WINDOW_AUTOSIZE); namedWindow("MOG2 Model",CV_WINDOW_AUTOSIZE); namedWindow("kKNNoutput Model", CV_WINDOW_AUTOSIZE); Mat kernel = getStructuringElement(MORPH_RECT,Size(3,3),Point(-1,-1)); //初始化BS Ptr<BackgroundSubtractor> pMOG2 = createBackgroundSubtractorMOG2(); Ptr<BackgroundSubtractor> pKNN = createBackgroundSubtractorKNN(); while (capture.read(frame)) { imshow("input video", frame); // MOG BS pMOG2->apply(frame, bsmaskMOG2); //形态学操作--开操作,去除小的噪声morphologyEx() morphologyEx(bsmaskMOG2, bsmaskMOG2, MORPH_OPEN, kernel, Point(-1, -1)); imshow("MOG2 Model", bsmaskMOG2); // KNN BS mask pKNN->apply(frame, bsmaskKNN); imshow("KNNoutput Model", bsmaskKNN); char c = waitKey(100); if (c == 27) { break; } } capture.release(); waitKey(0); return 0; }
python
#!/usr/bin/python3.6 # -*- coding: utf-8 -*- # @Time : 2020/11/17 19:06 # @Author : ptg # @Email : zhxwhchina@163.com # @File : 去背景.py # @Software: PyCharm import cv2 as cv import numpy as np from cv2 import cv2 image = cv2.imread("mabaoguo2.jpg",cv2.IMREAD_GRAYSCALE) binary = cv2.adaptiveThreshold(image,255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY_INV,25,15) se = cv2.getStructuringElement(cv2.MORPH_RECT,(1,1)) se = cv2.morphologyEx(se, cv2.MORPH_CLOSE, (2,2)) mask = cv2.dilate(binary,se) cv2.imshow("image",image) mask1 = cv2.bitwise_not(mask) binary =cv2.bitwise_and(image,mask) result = cv2.add(binary,mask1) cv2.imshow("reslut",result) cv2.imwrite("reslut00.jpg",result) cv2.waitKey(0) cv2.destroyAllWindows()
import cv2 import numpy as np #读入图像 video = cv2.VideoCapture("E:\\video.avi") videoIsOpen=video.isOpened print(videoIsOpen) width=int(video.get(cv2.CAP_PROP_FRAME_WIDTH))#宽度 height=int(video.get(cv2.CAP_PROP_FRAME_HEIGHT))#高度 fps=video.get(cv2.CAP_PROP_FPS)#获取帧率 print(fps,width,height) #创建窗口 cv2.namedWindow('MOG2') cv2.namedWindow('MOG22') cv2.namedWindow('input video') #cv2.namedWindow('KNN') bsmaskMOG2 = np.zeros([height,width],np.uint8) bsmaskKnn = np.zeros([height,width],np.uint8) #两种消除的方案 pMOG2 = cv2.createBackgroundSubtractorMOG2(detectShadows=True) PKNN = cv2.createBackgroundSubtractorKNN(detectShadows=True) #形态学处理 kernel = cv2.getStructuringElement(cv2.MORPH_RECT,(3, 3)) while videoIsOpen: (flag,frame)=video.read() if not flag: break cv2.imshow('input video',frame) # bsmaskKnn= PKNN.apply(frame) # cv2.imshow('KNN',bsmaskKnn) bsmaskMOG2 = pMOG2.apply(frame) cv2.imshow('MOG22',bsmaskMOG2) OPEND=cv2.morphologyEx(bsmaskMOG2,cv2.MORPH_OPEN,kernel) cv2.imshow('MOG2',OPEND) c = cv2.waitKey(40) if c==27: break video.release() cv2.waitKey(0)
以上就是OpenCV实现去除背景识别的方法总结的详细内容,更多关于OpenCV去除背景识别的资料请关注我们其它相关文章!
赞 (0)