how to detect car from a video
    3 次查看(过去 30 天)
  
       显示 更早的评论
    
foregroundDetector = vision.ForegroundDetector('NumGaussians', 3, ...
        'NumTrainingFrames', 50);
    V=VideoReader('video2.mp4');
    videoReader = vision.VideoFileReader('video2.mp4');
    for i = 1:150
        frame = step(videoReader); % read the next video frame
        foreground = step(foregroundDetector, frame);
       end
       imwrite(frame,'F:\referenceimage.jpg','jpg');
       videoPlayer = vision.VideoPlayer('Name', 'Detected Cars');
       videoPlayer.Position(3:4) = [650,400];  % window size: [width, height]
       se = strel('square', 3); % morphological filter for noise removal
       referenceimage=imread('F:\referenceimage.jpg'); 
       X=zeros(2,121);
       Y=zeros(2,121);
       Z=zeros;
       while ~isDone(videoReader)
           frame = step(videoReader); % read the next video frame
           % Detect the foreground in the current video frame
           foreground = step(foregroundDetector, frame);
           % Use morphological opening to remove noise in the foreground
           filteredForeground = imopen(foreground, se);
           %-----------------------SPEED ---------------------------%
           frame2=((im2double(frame))-(im2double(referenceimage)));
           frame1=im2bw(frame2,0.1); 
           [Labelimage]=bwlabeln(frame1); 
           stats=regionprops(Labelimage,'basic');
           BB=stats.BoundingBox;
           i=2; %fblasst for
           X(i)=BB(1);
           Y(i)=BB(2);
           Dist=((X(i)-X(i-1))^2+(Y(i)-Y(i-1))^2)^(1/2);
           Z(i)=Dist;
           M=median(Z);
           %disp(M);
           %clc;
           %disp('speed=')
           Speed=((M)*(120/8))/(4);
           %disp(Speed);
           %SPEED = M ???????????
           i = i + 1;
           SE = strel('disk',6); 
           frame3=imclose(frame1,SE); 
           step(videoReader); 
           pause(0.05);
              %if(i==121) end; ??
        %-----------------------SPEED ---------------------------%
        % Detect the connected components with the specified minimum area, and
        % compute their bounding boxes
        blobAnalysis = vision.BlobAnalysis('BoundingBoxOutputPort', true, ...
        'AreaOutputPort', true, 'CentroidOutputPort', true, ...
        'MinimumBlobArea', 150);
        %bbox = step(blobAnalysis, filteredForeground);
        [areas, centroids, bbox] = step(blobAnalysis, filteredForeground);
        % Draw bounding boxes around the detected cars
        result = insertShape(frame, 'Rectangle', bbox, 'Color', 'green');
        % Display the number of cars found in the video frame
        %ICI
        %disp(centroids); //show...
        %disp(size(centroids));//show....
        numCars = size(bbox, 1); %cars ...
        numCars_str = ['car number:', num2str(size(bbox, 1))];
        speed_str = [num2str(Speed),'KM/h'];
        result = insertText(result,[10,10],numCars_str, 'BoxOpacity', 1, ...
            'FontSize', 14);
        for i=1:size(bbox, 1)
        result = insertText(result,[centroids(i,1),centroids(i,2)], speed_str, 'BoxOpacity', 1, ...
            'FontSize', 20);
           end;
           step(videoPlayer, result);  % display the results
       end;
       release(videoReader); % close the video file
4 个评论
  Abhishek Kumar
    
 2020-12-11
				Hi Sandhya you can use "regionprops" function to get pixel values and several other parameters, if you are able to detect the car, in addition to link shared by KSSV, you can use the following link:
  Amirullah Bin Abdul Razak
 2023-1-7
				Hi, 
Appreciate if someone explain this part? Which 2 frame should i use?
frame2=((im2double(frame))-(im2double(referenceimage)));
Thank you,
Amir
回答(0 个)
另请参阅
类别
				在 Help Center 和 File Exchange 中查找有关 Tracking and Motion Estimation 的更多信息
			
	Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!



