Multiple Object Detection with Color Using OpenCV 

amazing! better than most object detection videos. thank you. :~) .

Great example! I had to modify a few lines for opencv2 v4 but works great!

For anybody who face the problem like mentioned by Arsapol and Farid Alijani. From what I found the main problem is the given multipleTrackingObject.cpp isn't complete. It lost 2 functions: CannyThreshold and trackFilteredObject. One note about trackFilteredObject is there should be 2 trackFilteredObject functions, one with 3 variables (this is the lost function) and the other one has 4 variables. this is the cause of those 2 appeared errors. I hope Ahmad Kaifi can help us to revise its multipleTrackingObject.cpp because from his video seems like he can run the program without any issue. Happy Coding^^

I really like the additional features over and above the work Kyle Hounslow did in the original (like the solid boundary element.) I do have a question though. It looks like you are using a different tool to generate the windows displayed in this youtube video. My console 32bit app has just plain windows, but your has some top bar options displayed. What tool are you using to generate those? Many thanks for posting both the video and your code sample on github.

Thanks your project is the best and helped me in my project as well... :D loved it

hello Ahmad Kaifi , already used OpenCV with Unity?

Very nice! What camera are you using? I'm just beginning with OpenCV and Python. I'm trying to identify a red laser dot but my webcam is too lousy...

Have you tried to using a TIME OF FLIGHT camera instead anytime? Have you done any similar work on Tof cameras? If yes, could you please share those videos too?

this very well , i have perform similar a project > thanks you so much

im just wondering. can it track certain shaped object? ex: any green color will not give you any output, but it will give you certain output, if it detects green colored square. Will it be possible?

I want to detect object with color, Distance upto 20 feet, with high speed, spherical shape, OCR etc.. there are different algorithms cover these features separately, but I want all these feature in my single programming code. So I need help to confirm or suggest way; 1 - Can I use single camera with for multiple algos? OR 2 - I need Separate camera for each Algo? 3 - If separate camera for each algo, then Is it possible for me to use multiple Algos in a single program? Any help/hint or clue will be appreciable.

Hi Ahmad, I'm Luyi who is your fan after watching your video "Multiple Object Detection". I use Open CV in Linux. Also, I don't use webcam but a video containing a lot of colourful balls. And the all codes has been compiled successfully. However, I meet with the following problems. First, in the video all the colourful balls can't be detected. Second, There is no name on the top of the balls in the video. And the balls couldn't be circled. By that I mean no thin border is drawn around the ball. In the video some noise places write the words "object". I don't what's wrong with your codes. I just make a change to the downloaded codes to use the video sources instead of the webcam. I don't what to do next. I notice that throughout the project the window3" After Morphological Operations" has not been displayed yet. why? Could you please give me some suggestions? Thank you very much.

does your code work on opencv 2.4.10? I don't understand why the "string" is not defined...

I always liked this demo. Did you do it as a school project or out of curiosity as part of a hobby. Did you want to apply the knowledge to a particular project? Are you still pursuing this area of interest?

Can you give tutorial detecting object and tracking object without color? I mean, detecting object with object form. I tried some tutorial object detecting with opencv but I can't get perfect detecting white object. So, I request to you to give a tutorial or you can give me private learning. Thank you. I hope you will help me.

Hi i am not able to track multiple objects of different colors (like in Video)can you guide me how to do that Thanks

I am getting errors. identifier "Scalar" is undefined and identifier "Mat" is undefined. I believe these have been depreciated or replaced in CV 3.2.0 or is that incorrect?

amazing keep it up

Hi. Is it possible to somehow recorded the trajectory of motion of the object?

Hello! I try to run your project but it does not recongnized colors of the objects. I have open 4 window at one i have video from the web cam but when i put colors object nothin happened

thank you so much, I run your code and camera data is always NULL. Could you please give me suggestion? Thanks

Muy interesante..seguire con la investigacion..gracias..

I want to count color object .Help me please ?

beautiful music

I stuck with the" trackFilteredObject(threshold,HSV,cameraFeed);" (in main) i don't know what should put the object in --> trackFilteredObject has 4 variables to put in right ?

Hey Dear.Akaifi Special thanks one question what's your compiler (IDE) in linux? pls answer...

excellent iam beginner but i have one question after i was detect object how can i take action

In Calibration mode, i had detected data of HSV then set in Object.cpp file but when i ran it again, it still couldnt detect that color( no contour, no label name). Did i miss some steps?

how about java program for detection color and puttext in open cv (eclipse) ? I've tried with a java program in open cv for color detection, but for putText any problem .. *give text on object May I know your email?

great work

I have written the code in python to find contours and draw them around the detected object but it is not working for outdoor images or cctv feeds of many people in single frame. Can you suggest me ways or some tutorial for the purpose?? Thank You!!

Which type of camera is used above?

in main() function when (CalibrationMode== true) you are using the following canny edge detection //the folowing for canny edge detec /// Create a matrix of the same type and size as src (for dst) dst.create(src.size(), src.type()); /// Convert the image to grayscale cvtColor(src, src_gray, CV_BGR2GRAY); /// Create a window namedWindow(window_name, CV_WINDOW_AUTOSIZE); /// Create a Trackbar for user to enter threshold createTrackbar("Min Threshold:", window_name, &lowThreshold, max_lowThreshold); /// Show the image trackFilteredObject(threshold, HSV, cameraFeed); my question is what is the use of "src_gray" you are not using it any where. Can u please explain?

Does anyone had problems with the drawObjects function? I'm getting an error saying there's no matching function, and I didn't actually understand how works the paremeters for this function, since in its declaration there's more parameters than where its being called. I appreciate any help.

Hi there, I am trying to implement the same thing, my problem at the moment is that once I've created the trackbars, they cannot move. I do not understand how to "on_trackbar" function works or where it came from, can you help?

hi.. this program is using c++. is there a way to translate it into python? have you used it with raspberry pi? thank you. :)

What is software which using on this video?

I am trying to make a self-driving car that would drive towards the balloons and pop them. I'd prefer to use a raspberry pi. Any suggestions?

Hello, i try the code. Why it error when run debug or release. Vector subscript out of range. Debug Assertion Fail? Using vs c++ 2010 and vs c++ 2013 all through this error.

Hello Ahmad can you tell me how to detect balls with same color????

AOA sir. how can I contact with you i need some guidance regarding this type of work that you showed in the video.

trackFilteredObject(???,threshold,HSV,cameraFeed); ???=What should i fill here?

Hey I don't know if you're still active here years later but could you please post the whole android studio project? Cos I have problems to make a full running project out of this

Hello, i am getting this error " fatal error C1083: Cannot open include file: 'cv.h': No such file or directory " . What should i do?

Hi, OpenCV Develops With....? for you project

is it possible to apply this on a desktop screen?

my friend, there's sth wrong in the code I guess: in the MultipleObjectTracking, I got these two errors: 1. CannyThreshold??? createTrackbar( "Min Threshold:", window_name, &lowThreshold, max_lowThreshold, ???CannyThreshold???????); 2. in the class trackedFilteredObject: the first element (Object) is missing for some reason: trackFilteredObject(-------------????, threshold,HSV,cameraFeed);

How do you get the HSV Max and min for a another color?

Hi! Does this work for pictures as well???

do you have .py file for the codes

May a start this with a makefile, if may, please send the code on makefile(I am russian)

sorry, in new version you change code (line 160) and need use drawObject(objects,cameraFeed,temp,contours,hierarchy); to compile, in my case I get error, when I try run .exe, MSVCP110D.dll and can´t fix this x(

Erros in CannyThreshold and trackFiltredObject Hi , i'm using opencv 3.0.0 and visual studio 2013 pro when i build my project i get the following errors CannyThreshold : undefined identifier trackFiltredObject : function does not take 3 arguments

Hi +Ahmad Kaifi you dont know how I run Multiple Object Detection in the terminal of linux or another person that know about this:(

which part is modified?

can it be ran in OpenCV3.0.0 ?

Hi, Great work. I am a newbie in this field and have a similar kind of project. Can you please share details and code for this design, if possible. Thanks

Hey, ahmad can you give the source code. Please it will help my research.

this code not work...why?

good job. may u share your codes for me ? please! i have a problem in this . thank u so much

You can send or share the code of this work, please, y need to do a similar work. thank you.

java!!

great work