Real-Time Object Tracking using YOLOv8 and DeepSORT | Vehicles Counting (Vehicles Entering& Leaving)

#yolo #yolov8 #objectdetection #objectracking #opencvpython #opencv #computervision #machinelearning #artificialintelligence #deepsort #multiobjecttracking#deeplearning
YOLOv8 and DeepSORT: A Powerful Combination for Object Tracking
Implement Multiple Object Tracking in Python with YOLOv8 and DeepSORT Tracking Algorithm.
-----------------------------------------------------------------------------------------------------------------------------
If you enjoyed this video, be sure to subscribe and press the 👍 button
🧑🏻‍💻 My AI and Computer Vision Courses⭐:
📙 YOLOv9: Learn Object Detection, Tracking with WebApps (13$):
www.udemy.com/course/yolov9-l...
📕 Learn LangChain: Build #22 LLM Apps using OpenAI & Llama 2 (13$):
www.udemy.com/course/learn-la...
📚 Computer Vision Web Development: YOLOv8 and TensorFlow.js (13$):
www.udemy.com/course/computer...
📕 Learn OpenCV: Build # 30 Apps with OpenCV, YOLOv8 & YOLO-NAS (13$):
www.udemy.com/course/learn-op...
📗 YOLO-NAS, OpenAI, SAM with WebApps using Flask and Streamlit (13$): www.udemy.com/course/yolo-nas...
📘 YOLO-NAS The Ultimate Course for Object Detection & Tracking (13$): www.udemy.com/course/yolo-nas...
📙 YOLOv8: Object Detection, Tracking & Web Apps in Python 2023 (13$) : www.udemy.com/course/yolov8-t...
📚 YOLOv7 YOLOv8 YOLO-NAS: Object Detection, Tracking & Web Apps in Python 2023 (13$): www.udemy.com/course/yolov7-o...
Follow Me:
LinkedIn: / muhammad-moin-7776751a0
GitHub: github.com/MuhammadMoinFaisal
Chat with us on Discord:
► / discord
For Consultation Call:
www.upwork.com/freelancers/~0...
--------------------------------------------------------------------------------------------------------------------------
Github Repo:
github.com/MuhammadMoinFaisal...
For Queries: You can comment here or email me at [e-mail]

[@muhammadmoinfaisal]

🧑🏻‍💻 My AI and Computer Vision Courses⭐: 📙 YOLOv9: Learn Object Detection, Tracking with WebApps (13$): www.udemy.com/course/yolov9-learn-object-detection-tracking-with-webapps/?couponCode=MAY13DOLLAR 📕 Learn LangChain: Build #22 LLM Apps using OpenAI & Llama 2 (13$): www.udemy.com/course/learn-langchain-build-12-llm-apps-using-openai-llama-2/?couponCode=MAY13DOLLAR 📚 Computer Vision Web Development: YOLOv8 and TensorFlow.js (13$): www.udemy.com/course/computer-vision-web-development/?couponCode=MAY13DOLLAR 📕 Learn OpenCV: Build # 30 Apps with OpenCV, YOLOv8 & YOLO-NAS (13$): www.udemy.com/course/learn-opencv-build-30-apps-with-opencv-yolov8-yolo-nas/?couponCode=MAY13DOLLAR 📗 YOLO-NAS, OpenAI, SAM with WebApps using Flask and Streamlit (13$): www.udemy.com/course/yolo-nas-object-detection-tracking-web-app-in-python-2023/?couponCode=MAY13DOLLAR 📘 YOLO-NAS The Ultimate Course for Object Detection & Tracking (13$): www.udemy.com/course/yolo-nas-the-ultimate-course-for-object-detection-tracking/?couponCode=MAY13DOLLAR 📙 YOLOv8: Object Detection, Tracking & Web Apps in Python 2023 (13$) : www.udemy.com/course/yolov8-the-ultimate-course-for-object-detection-tracking/?couponCode=MAY13DOLLAR 📚 YOLOv7 YOLOv8 YOLO-NAS: Object Detection, Tracking & Web Apps in Python 2023 (13$): www.udemy.com/course/yolov7-object-detection-tracking-with-web-app-development/?couponCode=MAY13DOLLAR

[@wositangaiaiai6416]

Hello, Muhammad Moin，I am a student of a Chinese university. Your Yolov8+deepsort project has greatly helped me with my graduation thesis. Thank you for your contribution in this field!

[@muhammadmoinfaisal]

Hi @Wositang, very happy to hear, and if you need any further help do let me know All the best for your future endeavors Thanks

[@CoderInfo-wh1zq]

ug or pg in chinese university

[@CoderInfo-wh1zq]

dear wositang aiaiai can you talk to me

[@sahalbelam1134]

what are y doing right now sir\ \

[@elianabboud8721]

Can you send me your thesis or the correct code

[@ainulhakimbinfizalsabillah2107]

Hi, thanks for the great video. i have an issue with pytorch. the code only allow using torch 2.0.1 cpu. and if i install torch 2.0.1 cu118 which is cuda version. it encounter error. is the code didnt implement cuda backend?

[@davidhasiansimatupang9869]

thanks for the tutorial, its very helpful, but can you make a tutorial video for training custom dataset for detecting and counting vehicle and than the result is also save in csv file

[@fayezalhussein7115]

Great work, thank you so much

[@muhammadmoinfaisal]

Glad you liked it!

[@RajendraKusuluri]

hi how do we modify the code to only track person in the video file is there any option track only person class like yolo v7 had --classes 0 for the person

[@user-wb5im5kp8f]

Such an amazing video! What should I do if I want to track an ID and know how long it has been present？Just like issuing some warnings to the terminal or server after the ID appears for more than 60 seconds.

[@Simon-tw1vf]

Hello, I get following error when trying to install ultralytics. Can you help? (venv) D:\DeepSORT\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect>pip install ultralytics ERROR: Could not find a version that satisfies the requirement ultralytics (from versions: none) ERROR: No matching distribution found for ultralytics

[@jiaojiaolou]

Hello, Muhammad Moin,the tracking folder inside the link cannot be found, can I ask you to help me?

[@yolov-ib5te]

Great work as always

[@muhammadmoinfaisal]

Thank you! Cheers!

[@yolov-ib5te]

Looking forward for more videos on YOLOv8

[@muhammadmoinfaisal]

Thanks a lot

[@user-ts6rt2zr2t]

Hello, I am a Chinese student doing design. Your work has helped me a lot. I would like to ask if this can run on Mac?

[@anime_on_data7594]

i have seen a major chnages in the repo and unable to relate to this video, Moin can you specify some?

[@journeyingpalates]

Hello Sir, while. training the model, the value for validation loss, is showing as 0,infinity and 0 for box, classification and dlf loss, how can I get accurate values as training loss

[@sriram-jp4xj]

the video is really helpful. a tutorial video deep sort would be really helpful.

[@muhammadmoinfaisal]

Hi @sri ram, yes will be uploading in next few days

[@uzairalikhan8557]

Hello brother! Can I use this for particle tracking by some modification?

[@pawansharma-mi6ce]

make video on vehicle on the basic of direction in unsignalized interasection

[@SunEside]

How can I count cars with two diagonal lines not horizontal lines?? please teach my how to do this

[@Zinab8850]

Can I count each vehicle time it took to pass the road?

[@mishitpatel]

Thanks for the video. I have seen this exact example, many places. May I know the original contributors for this codebase? Appreciate your credits.

[@Nico-hf6hw]

You are an absolute legend! you have a new sub. Can this be done in C++?

[@muhammadmoinfaisal]

Hi @Nico, yes it can be done but you have to make certain changes for this

[@vishnum7985]

Really helpful content. Can you tell me the best tracking method that I can use in current scenario. I wanted to know if ByteTrack is better than deepsort or vise versa

[@muhammadmoinfaisal]

Hi @Vishnu, from my analysis DeepSORT gives best results, but i am also working on creating a tutorial on ByteTrack Tracking as well and will be published soon

[@ishankamadhuranga3729]

Can you show how to predict movement of object

[@KangReposting]

hello sir, how do you get the Vehicle Count results to be output as a csv file?

[@sankalpmishraOFFICIAL]

Got the answer ?

[@juliangalindo3440]

After processing each frame you can update a dictionary or something like that registering unique ids (initializing a set object for example) and make a pandas dataframe to dump the results to a csv

[@ekincanuzun4038]

How can we evaluate this model?

[@amishachauhan375]

​@Muhammad Moin how can we evaluate the performance of this model. Can you help with the same?

[@sourabjain3947]

hi, were you able to figure this out? if yes, could you please share?

[@aleshanmaistry1367]

Hi I cloned the repo but i do not see the tracking and tracking_vehicle_count files in the repo even the repo instructions have changed can you please assist

[@muhammadmoinfaisal]

HI @Aleshan, yes the instructions have changed a bit, please review the steps to run code in the github repo to run the tracking and tracking_vehicle_counting

[@chaimacham9414]

please how can i use this with yolov7 can you explain please

[@aayushgupta1034]

kindly make a video on how to train multi object tracking model on visdrone dataset.

[@muhammadmoinfaisal]

Sure I will make a video on it

[@angshumanpandit367]

How can i get txt file of tracked objects

[@roobeshbalaji9282]

is it possible to use a webcam and have a live feed as the source instead of a video clip?

[@muhammadmoinfaisal]

Yes, just remove the video path and add 0 over there to run it on Live Webcam

[@roobeshbalaji9282]

@@muhammadmoinfaisal perfect. Also how do I make it run on my GPU? It seems to be running on my CPU. Is that possible?

[@alekfgh9734]

Thank you very much, it helped me a lot, how can I do so that it does not use the CPU and only the GPU?

[@muhammadmoinfaisal]

Hi Alek, Welcome, If you have GPU in your system then you can simply run the script and it will use GPU, but if you dont have GPU based system, try Google Colab it offers free GPU and colab file for this implementation is uploaded in the github repo. Do check it out and let me know if there is any issue

[@alekfgh9734]

@@muhammadmoinfaisal I implemented it but one part runs on the GPU and another on the CPU and consumes a lot of CPU

[@alekfgh9734]

@@muhammadmoinfaisal How do I disable the training function? I don't want the video of the results to be saved

[@joart333]

Superb Sir But audio is only hearable from left side.

[@muhammadmoinfaisal]

Hi @Joart3, i will make sure to improve it in upcoming videos Thanks a lot

[@akifakbulut765]

There is no tracking file in the yolo folder on your github page, where can I access this file.

[@muhammadmoinfaisal]

Hi @Akif, the file names have been updated can you please check the steps to run code in the github repo, it will help you Thanks

[@-----zg1qc56]

@@muhammadmoinfaisal could you please say the name because i checked all those folders, you play a game

[@aadrikabhargava5540]

Even after running the final command in your repository I can't get the count of vehicles , I only get the one in the previous step

[@muhammadmoinfaisal]

Hi Aadrika, Please check the colab file, its a single click solution, and did you download the updated predict.py file from the Google drive for vehicle counting?

[@aadrikabhargava5540]

Hey got it , also How can I modify the command you have given and run it on gpu ?

[@nivitusfernandez3574]

Really appreciated sir, After using run command I have an issue like "[2023-01-14 22:56:37,481][root.tracker][INFO] - Loading weights from deep_sort_pytorch/deep_sort/deep/checkpoint/ckpt.t7... Done!" Can't see any inference video!

[@muhammadmoinfaisal]

Hi @Nvitus are you facing this issue on Windows or while running Google colab script

[@nivitusfernandez3574]

@@muhammadmoinfaisal Sir actually I'm using Ubuntu 20.04. I had an issue while running Ubuntu local machine

[@Nico-hf6hw]

I got the same issue!

[@muhammadmoinfaisal]

Hi @Nivitus I have updated the github repository please check it now you will not face this issue with you try on windows Linux it will work fine

[@nivitusfernandez3574]

@@muhammadmoinfaisal Thank you so much! It's working

[@mosesmaned2151]

please is it necessary to import all of the files please

[@MohamedAdel-ln5lt]

i am student and my graduating project is on detection and tracking vehicles the detecting is done but the tracking is my problem can you me on how to just integrate this program into the GUI i have only 5 days

[@muhammadmoinfaisal]

You can check the YOLO-NAS Video Series and check the video tutorials on how to Create a Webapp using Streamlit and Flask by integrating YOLO-NAS with Flask and Streamlit

[@user-zq7zg3rc1e]

my friend ,why is this a problem File "predict.py", line 22, in import ultralytics.yolo.engine.predictor ModuleNotFoundError: No module named 'ultralytics.yolo' can't find 'ultralytics.yolo' why?

[@salmabenmustapha9413]

thank you very much !!! can you pleaseee make a video on how to adapt this project to perform custom object tracking? (detect specific objects rather than the 80 objects detected by YOLO for exemple , or detect only people and track them ) ? that would be very helpful as it's the issue i am currently facing!

[@muhammadmoinfaisal]

Hi @Salma, i will be uplaoding a video tutorial on it today please check it

[@salmabenmustapha9413]

@@muhammadmoinfaisal thank you !

@@muhammadmoinfaisal I can't wait for the video!

[@vitormatheus8112]

Boa tarde, o meu está dando erro no "hydra" No module named hydra. como faço pra ajustar?

[@muhammadmoinfaisal]

Hi Vitor you are getting this error because you have skipped this step pip install -e '. [dev]' Please install dependencies

[@vitormatheus8112]

​@@muhammadmoinfaisal [Errno 2] No such file or directory - Hello Muhammad is giving this error, even though I follow all the step by step of your tutorial, can you help me please

[@elianabboud8721]

I am facing an error while run. Does anyone have the correct code?

[@user-du2kz3ks8g]

After cloning the project i there are missing some files like tracking.py. what should i do?

[@zzzzzyh8794]

i have th same question as you

[@nivitusfernandez3574]

I couldn't find Tracking_Vehicle_counting.py, Where it's, Can you please help out this?

[@muhammadmoinfaisal]

Hi @Nvitus, please review the read me file in the github repo the step by step instructions to run the script is provided. In the updated github repo the tracking_vehicle_counting.py file is replaced by predict.py Thanks

[@rorokemuningsari3572]

hello sir, what if I want to use the model and class that I created myself?

[@muhammadmoinfaisal]

For this you will have to train the YOLOv8 model on the custom dataset first for a specific class

[@rorokemuningsari3572]

@@muhammadmoinfaisal I have performed training using two classes (motorcycle and car), which resulted in the best.pt and last.pt models, as well as the argm.yaml file. So, how can I modify the model and classes in the code? I'm sorry because I'm a beginner

[@nguyenthevinh8099]

Thank you verry much. Now, i don't want use: python tracking_vehicle_counting.py model=yolov8n.pt source="test.mp4" show=True. I want use API python , how to use?

[@muhammadmoinfaisal]

Hi @Nguyen, I will be making a separate video on it in next week

[@bmumut]

The more I solve it, the more errors I encounter. it does not work

[@aadrikabhargava5540]

How to run this on gpu ?

[@muhammadmoinfaisal]

Hi @Aadrika bhargava, to run on GPU please follow the same steps as you are following for CPU, but please make sure that you have Torch with GPU enabled

[@muhammadmoinfaisal]

Some fellows have highlighted they are facing issue while running the script I have updated the github repo now it will be working fine either you try in windows, Linux, Mac Github Repo: github.com/MuhammadMoinFaisal/YOLOv8-DeepSORT-Object-Tracking Thanks

[@nabinsharma7377]

Hey there is no track file?

[@muhammadmoinfaisal]

Hi @Nabin Sharma, please check the github repo, and please have a look at the steps to run code, in the steps to run code, the detail instructions have been provided how to implement tracking plus tracking and vehicle counting as well. If you have any issue let me know github.com/MuhammadMoinFaisal/YOLOv8-DeepSORT-Object-Tracking Thanks

[@darn2926]

Hello @Muhammed Moin, How can i place multiple lines if it where for a junction from 4 different direction. Thank you in advance, and sir I have sent you an email thank you in advance.

[@muhammadmoinfaisal]

Hi Darn, I might have missed your email, can you please resent it and tutorial for 4 different direction is coming by the end of this week or in next week start, if you have any video on which you want me to test, please send me on the email. [e-mail]

[@totuan4276]

YOLOv8s summary (fused): 168 layers, 11156544 parameters, 0 gradients, 28.6 GFLOPs Error executing job with overrides: ['model=yolov8s.pt', 'source=test3.mp4', 'show=True'] Traceback (most recent call last): File "/Users/Mutsu/Desktop/yolov8/YOLOv8-DeepSORT-Object-Tracking-main/ultralytics/yolo/v8/detect/predict.py", line 315, in predict predictor() File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context return func(*args, **kwargs) File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/ultralytics/yolo/engine/predictor.py", line 158, in __call__ return list(self.stream_inference(source, model)) # merge list of Result into one File "/Library/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/ultralytics/yolo/engine/predictor.py", line 199, in stream_inference self.results = self.postprocess(preds, im, im0s, self.classes) TypeError: postprocess() takes 4 positional arguments but 5 were given Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace. i got the error. can you fix it for me, please. Thank you, Mohammad.

[@muhammadmoinfaisal]

hi To Tuan did you follow all the steps provided in the github repo I just checked it on my system it's working fine plus you can also check the colab file as well it is a single click solution as well Thanks

[@amishachauhan375]

Hey I have same issue..were you able to resolve it?

[@simba12138]

i tried to use my own trained .pt file to perform tracking. i solved one BUG with pip install ultralytics==8.0.0 but it will cause a new problem which is "ModuleNotFoundError: No module named 'ultralytics.nn.modules.conv'; 'ultralytics.nn.modules' is not a package; Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace."

[@Nico-hf6hw]

Hello thanks allot for this but im getting an issue while Loading the video: This what I wrote on the cmd console: python tracking.py model=yolov8l.pt source="test3.mp4" show=True This is what I got: Loading weights from deep_sort_pytorch/deep_sort/deep/checkpoint/ckpt.t7... Done!

[@muhammadmoinfaisal]

Hi @Nico I have updated the github repository please check it now you will not face this issue with you try on windows Linux it will work fine

[@Nico-hf6hw]

@@muhammadmoinfaisal Hey I tried to run it. But there are som other issues. PS C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect> python predict.py model=yolov8n.pt source="test3.mp4" show=True [2023-01-17 20:33:55,262][root.tracker][INFO] - Loading weights from deep_sort_pytorch/deep_sort/deep/checkpoint/ckpt.t7... Done! Ultralytics YOLOv8.0.3 Python-3.9.15 torch-1.13.1+cpu CPU Fusing layers... YOLOv8n summary: 168 layers, 3151904 parameters, 0 gradients, 8.7 GFLOPs Error executing job with overrides: ['model=yolov8n.pt', 'source=test3.mp4', 'show=True'] Traceback (most recent call last): File "C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect\predict.py", line 252, in predict predictor() File "C:\Users ico_\anaconda3\envs\SecureVision\lib\site-packages\torch\autograd\grad_mode.py", line 27, in decorate_context return func(*args, **kwargs) File "c:\users ico_\documents\yolov8-deepsort-object-tracking\ultralytics\yolo\engine\predictor.py", line 189, in __call__ s += self.write_results(i, preds, (p, im, im0s)) File "C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect\predict.py", line 234, in write_results outputs = deepsort.update(xywhs, confss, oids, im0) File "C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect\deep_sort_pytorch\deep_sort\deep_sort.py", line 31, in update detections = [Detection(bbox_tlwh[i], conf, features[i],oid) for i, (conf,oid) in enumerate(zip(confidences,oids)) if conf > self.min_confidence] File "C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect\deep_sort_pytorch\deep_sort\deep_sort.py", line 31, in detections = [Detection(bbox_tlwh[i], conf, features[i],oid) for i, (conf,oid) in enumerate(zip(confidences,oids)) if conf > self.min_confidence] File "C:\Users ico_\Documents\YOLOv8-DeepSORT-Object-Tracking\ultralytics\yolo\v8\detect\deep_sort_pytorch\deep_sort\sort\detection.py", line 30, in __init__ self.tlwh = np.asarray(tlwh, dtype=np.float) File "C:\Users ico_\anaconda3\envs\SecureVision\lib\site-packages umpy\__init__.py", line 285, in __getattr__ raise AttributeError("module {!r} has no attribute " AttributeError: module 'numpy' has no attribute 'float' Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

[@nathapongofficial5837]

​@@Nico-hf6hw same problem

[@muhammadmoinfaisal]

@@nathapongofficial5837 @Nico, I have just checked it, its working completely fine and didnot face any issue in my Local IDE, plus using the same steps mentioned in the github repo i implemented in Google Colab and it worked completely fine, the Google Colab File is uploaded in the Github repo, you can check and run it on you side, if you face any issue please let me know i am available

[@muhammadmoinfaisal]

HI @Nico @Nathapong Official, we have updated the github repo please check it now you will not face this issue Thanks

[@dasigiraghu640]

python predict.py model=yolov8l.pt source="test3.mp4" show=True Traceback (most recent call last): File "predict.py", line 14, in from ultralytics.yolo.utils import DEFAULT_CONFIG, ROOT, ops ImportError: cannot import name 'DEFAULT_CONFIG' from 'ultralytics.yolo.utils' (C:\Users aghu.dasigi\anaconda3\envs\YOLOv8-DeepSORT-Object-Tracking\lib\site-packages \ultralytics\yolo\utils\__init__.py) Sentry is attempting to send 2 pending error messages Waiting up to 2 seconds unable to find DEFAULT_CONFIG

[@muhammadmoinfaisal]

Hi @Dasgi, I have just run the script i didnot get this issue, can you please rerun or check the colab file again and let me know Thanks

[@dasigiraghu640]

@@muhammadmoinfaisal Hi , this error was resolved when i used ultralytics from git hub repo instead of pip..the name was different ....Thank you

[@oguzcanbekar2536]

@@dasigiraghu640 I also face the same problem how did you solve it can you explain please?

[@dasigiraghu640]

@@oguzcanbekar2536 don't do pip install ultralytics... That's all

[@amarsaish5931]

@@dasigiraghu640 I'm getting the same error there was no DEFAULT_CONFIG , in the file also there wasnt mentioned any class about DEFAULT CONFIG

[@taharezaei856]

Hello, Muhammad Moin， My name is Taha, I am a student at Exeter University, UK. I have a question about Yolo 8 can I have your email?

[@muhammadmoinfaisal]

Sure [e-mail]