Tracking balls in a video and find their specific locations when are leaving a specific point.

Hi everyone. I need your help. It concerns a project that i need to do. Does anyone knows how i am tracking a balls in a video and find their specific location and time that left from a circle shape? I keep trying find something similar but i couldn't find something. Any help will be very appreciated. Thanks a lot.

Answer 1

[Image Analyst]

My attached demo where I track a green Sharpie marker is similar.

[ZENONAS ZENIOU]

That was very helpful thanks a lot. Do you know the coding that i need to add to track balls?

[Image Analyst]

No. For just one blob, you simply just index the location. But it you have multiple blobs entering and leaving the field of view, and overlapping each other, it becomes quite complicated and I've never had to do that so I don't have any code for it.

[ZENONAS ZENIOU]

Hello again sir. The code was good but i have some questions and i will be very pleasant if you help me. Basically when i am trying to track one ball the diagram which shows the coordinated is not appear and doesnt show the coordinates and i dont know why. Moreover is there any code that i need to add in order to get all the coordinates and then connecting all of them in order to show the path that the ball followed? Thanks a lot.

[Image Analyst]

You need to index your coordinates into an array and then plot the whole array each time instead of just the one centroid.

As far as coordinates not showing up, I have no idea unless I can run the code. Perhaps you're not doing your segmentation properly. Try it with just one image and see if your binary image is properly created.

[ZENONAS ZENIOU]

I use that code that you send me last time. Basically i need to track a moving marble and the coordinates are not showing up. Do you know where maybe there is a mistake?

%% Demo to track green color. Finds and annotates centroid and bounding box of green blobs. % Modify thresholds to detect different colors.

clc; % Clear the command window. close all; % Close all figures (except those of imtool.) imtool close all; % Close all imtool figures if you have the Image Processing Toolbox. clear; % Erase all existing variables. Or clearvars if you want. workspace; % Make sure the workspace panel is showing. format long g; format compact; fontSize = 20;

%% Specify input video file name. folder = pwd; % fullFileName = 'rhinos.avi'; % fullFileName = 'traffic.avi'; baseFileName = 'IMG_1213.mp4'; fullFileName = fullfile(folder, baseFileName);

%% Check if the video file actually exists in the current folder or on the search path. if ~exist(fullFileName, 'file') % File doesn't exist -- didn't find it there. Check the search path for it. fullFileNameOnSearchPath = baseFileName; % No path this time. if ~exist(fullFileNameOnSearchPath, 'file') % Still didn't find it. Alert user. errorMessage = sprintf('Error: %s does not exist in the search path folders.', fullFileName); uiwait(warndlg(errorMessage)); return; end end

%% Instantiate a video reader object for this video. videoObject = VideoReader(baseFileName);

%% Setup other parameters numberOfFrames = videoObject.NumberOfFrame;

%% Set HSV thresholds for the green sharpie in the demo video. % Modify the thresholds to detect different colors. hThresholds = [0.2, 0.44]; sThresholds = [0.2, 1.0]; vThresholds = [20, 125]; %% Read one frame at a time, and find specified color. for k = 1 : numberOfFrames % Read one frame thisFrame=read(videoObject,k); hImage=subplot(3, 4, 1); % Display it. imshow(thisFrame); axis on; caption = sprintf('Original RGB image, frame #%d 0f %d', k, numberOfFrames); title(caption, 'FontSize', fontSize); drawnow; hsv = rgb2hsv(double(thisFrame)); hue=hsv(:,:,1); sat=hsv(:,:,2); val=hsv(:,:,3);

subplot(3, 4, 2); imshow(hue, []); impixelinfo(); axis on; title('Hue', 'FontSize', fontSize); subplot(3, 4, 3); imshow(sat, []); axis on; title('Saturation', 'FontSize', fontSize); subplot(3, 4, 4); imshow(val, []); axis on; title('Value', 'FontSize', fontSize); if k == 1 % Enlarge figure to full screen. set(gcf, 'Units', 'Normalized', 'OuterPosition', [0 0 1 1]); %Give a name to the title bar. set(gcf, 'Name', 'Demo by ImageAnalyst', 'NumberTitle', 'Off') hCheckbox = uicontrol('Style','checkbox',... 'Units', 'Normalized',... 'String', 'Finish Now',... 'Value',0,'Position', [.2 .96 .4 .05], ... 'FontSize', 14); end

%Compute histograms for H, S, and V channels %Let's compute and display the histogram. [pixelCount, grayLevels] = imhist(hue); subplot(3, 4, 6); bar(grayLevels, pixelCount); grid on; %title('Histogram of hue image', 'FontSize', fontSize); xlim([0 grayLevels(end)]); % Scale x axis manually. %Let's compute and display the histogram. [pixelCount, grayLevels] = imhist(sat); subplot(3, 4, 7); bar(grayLevels, pixelCount); grid on; title('Histogram of sat image', 'FontSize', fontSize); xlim([0 grayLevels(end)]); % Scale x axis manually. % Let's compute and display the Value histogram. [pixelCount, grayLevels] = imhist(uint8(val)); subplot(3, 4, 8); bar(grayLevels, pixelCount); grid on; %title('Histogram of val image', 'FontSize', fontSize); xlim([0 grayLevels(end)]); % Scale x axis manually. binaryH = hue >= hThresholds(1) & hue <= hThresholds(2); binaryS = sat >= sThresholds(1) & sat <= sThresholds(2); binaryV = val >= vThresholds(1) & val <= vThresholds(2); subplot(3, 4, 10); imshow(binaryH, []); axis on; title('Hue Mask', 'FontSize', fontSize); subplot(3, 4, 11); imshow(binaryS, []); axis on; title('Saturation Mask', 'FontSize', fontSize); subplot(3, 4, 12); imshow(binaryV, []); axis on; title('Value Mask', 'FontSize', fontSize); % Overall color mask is the AND of all the masks. coloredMask = binaryH & binaryS & binaryV; %Filter out small blobs. coloredMask = bwareaopen(coloredMask, 500); % Fill holes coloredMask = imfill(coloredMask, 'holes'); subplot(3, 4, 9); imshow(coloredMask, []); axis on; title('Colored Blob Mask', 'FontSize', fontSize); drawnow; [labeledImage, numberOfRegions] = bwlabel(coloredMask); if numberOfRegions >= 1 stats = regionprops(labeledImage, 'BoundingBox', 'Centroid'); % Delete old texts and rectangles if exist('hRect', 'var') delete(hRect); end if exist('hText', 'var') delete(hText); end % Display the original image again. subplot(3, 4, 5); % Switch to original image. hImage=subplot(3, 4, 5); imshow(thisFrame); axis on; hold on; caption = sprintf('%d blobs found in frame #%d 0f %d', numberOfRegions, k, numberOfFrames); title(caption, 'FontSize', fontSize); drawnow; %This is a loop to bound the colored objects in a rectangular box. for r = 1 : numberOfRegions % Find location for this blob. thisBB = stats(r).BoundingBox; thisCentroid = stats(r).Centroid; hRect(r) = rectangle('Position', thisBB, 'EdgeColor', 'r', 'LineWidth', 2); hSpot = plot(thisCentroid(1), thisCentroid(2), 'y+', 'MarkerSize', 10, 'LineWidth', 1) hText(r) = text(thisBB(1), thisBB(2)-20, strcat('X: ', num2str(round(thisCentroid(1))), ' Y: ', num2str(round(thisCentroid(2))))); set(hText(r), 'FontName', 'Arial', 'FontWeight', 'bold', 'FontSize', 12, 'Color', 'yellow'); end hold off drawnow; end % See if they want to bail out if get(hCheckbox, 'Value') % Finish now checkbox is checked. msgbox('Done with demo.'); return; end end msgbox('Done with demo.');

[Image Analyst]

Read this, but better yet, just attach your code and video file with the paper clip icon. It should be simple for you to adapt it to find red instead of green - you probably just forgot to change the thresholds. If you don't change those of course, it will continue to find green.

[Image Analyst]

Do you want to know when the centroid of the ball leaves the big circle, or when any part of the perimeter of the ball leaves the big circle?

Anyway, I looked at the video and it looks like my suggestion in one of your 3 other duplicate questions should work. Just use my video reading program, and then simplify it. Like make the background the average of the first 20 frame or so instead an adaptive one from the past few frames. Then subtract the frame from the background and see where the difference is more than some amount. Then find the boundaries and see where any of those points are within a radius of the center of the big circle. Not hard. I'm sure you can do it.

Answer 2

[ZENONAS ZENIOU]

Basically is the code that you send me with the green sharpie. So now basically i need to track one ball every time.

Answer 3

[ZENONAS ZENIOU]

I have attached a file with the code. What basically i need to do is to saving the positions of each blob and the connect all the points and represent the path that the blob followed. Also imagine that in a video there is a circular shape where all the balls are moving through. What i need to do basically is to know where the blobs (ball) exiting that circular shape. Thanks a lot

[Image Analyst]

You posted this as an answer to your original question instead of as a follow up comment to me. So I'm assuming that your code that you attached here solves your problem and lets you "know where the blobs (ball) exiting that circular shape." Glad you got it working.

[ZENONAS ZENIOU]

I am really sorry. can you please review the code and if you can help me do the following?. What basically i need to do is to saving the positions of each blob and the connect all the points and represent the path that the blob followed. Also imagine that in a video there is a circular shape where all the balls are moving through. What i need to do basically is to know where the blobs (ball) exiting that circular shape. Thanks a lot

[Image Analyst]

No, I can't run it unless you also give a video that has the balls in it.

[ZENONAS ZENIOU]

How can i upload it because is larger than 5kb

[Image Analyst]

Crop out a small part of it. Anyway, you can do more than 5 kb. The file I uploaded was 87 kb and I upload images bigger than that all the time.

[ZENONAS ZENIOU]

I cut a little bit of the actual video but this is the concept. What basically i need to do is to saving the positions of each blob and the connect all the points and represent the path that the blob followed. Also imagine that in a video there is a circular shape where all the balls are moving through. What i need to do basically is to know where the blobs (ball) exiting that circular shape. Thanks a lot.

[Image Analyst]

DO you want to know when the centroid of the ball leaves the big circle, or when any part of the perimeter of the ball leaves the big circle?

When I try to play the file in Windows Media Player it just shuts down immediately. So I can't see this file. Does it work with MATLAB?

Answer 4

[Image Analyst]

Try the attached script.

[ZENONAS ZENIOU]

This is very helpful.Thanks a lot. Is there any way that i can know the coordinates where the ball exiting that circular shape?

[Image Analyst]

Z's "Answer" moved here because it's NOT an answer to the originally posted question, but actually a comment to me. Z, please don't post follow-up questions as new, independent answers to your original question. Use the Comment button to do that.

Sorry again. What basically i need to add to the code is the coordinates of each blob and then knowing where does the ball exiting the circular shape. thanks again for your help. Highly appreciated

[Image Analyst]

Look at the line

distance(frame) = sqrt((centerX - props(1).Centroid(1)).^2 + (centerY - props(1).Centroid(2)).^2);

You can see the centroid there. Just save it

centroidX = props(1).Centroid(1);
centroidY = props(1).Centroid(2);

[ZENONAS ZENIOU]

I did this but i can't see the coordinates. What i want if is possible is showing the coordinates of the blob while is moving like the green sharpie exercise. Can you suggest me how to find the coordinates where the ball leaving the circular shape? Thanks a lot again

[Image Analyst]

There is a very useful function called plot() that you should learn about. Just use that.

hold on
plot(centroidX, centroidY, 'r+', 'MarkerSize', 15, 'LineWidth', 2);

Look up plot in the help and explore its different options.

[ZENONAS ZENIOU]

Yeah i did the but nothing appeared. The think is that i want the locations where the balls exiting (leaving) the circular shape? Do you have any suggestions what i need to add to the coding in order to get those locations?

[Image Analyst]

No, because that should work. Perhaps something immediately clears it, like you're displaying another image? I put up outline masks over live video all the time and it's not a problem. Put a breakpoint on that plot call and then execute just that line and see if it puts up the red cross. If it does, then step through to find out what line of code makes it go away.

[ZENONAS ZENIOU]

The red cross appears. I just want to see the coordinates where the balls exiting the circle. For the time where not balls are in the frames this is shown up: Processed frame 16 of 230 Area of frame 16 = 17. And when the blob appears this is shown: Area of frame 53 = 6986. distance from circle center for frame 53 = 4.550689e+02. What basically i need to do as a final result of this code is to plot a histogram and showing from which site of the circular shape did the balls exited the circular shape. So i thing i need to know the location of the blob when exiting the circle or at least the last position of the ball inside the circle before exiting the circle? Thanks again for your help