Track & Count Objects and Speed Estimation using YOLOv8 ByteTrack & Supervision

hey viewers welcome back P channel here we share ground break example of speed estimation and L using object dation model and by track and supervision Library let's start welcome back to B channel in this video you will learn how to uh identify and also exem speed of the vehicle deduction and also tracking the speed so let's going start from from the sketch first of all you can see uh first I'm run at app.py it's demo for you and uh you will see we will track the object and here steam uh estimation speed so it's uh in front of you we will going to exploring multiple video on the multiple videos and we will uh Implement uh and changing some points in front of you uh in supervision uh Library so let's start from sketch first you need to understand how to code work okay so super Vision roof flow one example is available okay here the roboff flow library of the supervision and I am going to uh supervision examples okay so here and I'm going back and uh you see the examples okay so speed estimation here you can see the one example available and uh majority code is bike track get from here and uh man code is available in front of you here so it's very I'm implement it's going to easy and run it just python script name and you will get the result in within seconds so uh you see that this example speed station analy uh using varation object models bike track okay and a simple effective online multiple tracks and front of you you can see here supervision package multiple tracks will be DED andom Etc and uh we will going to uh exploring uh step by step okay so first you need to install this library and uh it's very simple and easy to install let's start from sketch okay just copy from here okay this library and I am close the terminal okay in front of you you can see you just open your CMD command and uh uh CMD and uh if you make user you just open the terminal if you uh just like a Linux user you just open your uh Linux terminal so we will open the our Mech terminal and pip install supervision library in front of you uh this Library will be done I am already installed this packet okay and uh I am minimize browser and uh I open it my in front of you this file okay so steam Mission speed file will be open and uh you can see here the one folder and uh uh here the one file okay so uh I am download it in front of you one simple example from internet and we will uh easily uh uh detect the all videos okay not one video all video so so in front of you you can see this example and and uh here this example and also last one it's same but the processing uh step stepping is slow slow so we will uh uh open the a file first of all okay so you need to understand this code and I am also mention the requirement files so here just pip install and uh requirement file name okay if you want to change this file name you can easily open it and you can see I'm just write the name okay and uh in front of you already install all dependency and now you need to understand this code first of all what will be changing in uh we achieve in this code and you can see we import the all libraries and uh you will see we will put the here report Li Library also and get get the video set up okay we will open the one video file mp4 file otherwise you will also open your webcam just replace the your zero and one one is the your internal SEC 0 one is your internal second one is the external webcam setup okay this code is get from the myet account so second part is the saving the video here two lines is saving the video third one you see the here this code okay this code will be mentioned the getting color of each track if you want to change color uh you will adjust and change the uh uh just like you separated adjust the uh here numbers okay and uh mention the your color okay after you can see we load the model we load the model ultris and Nano and now you will see the we are uh hide the polygon uh method and uh remove the track okay but in front of you we will put the poly on for for points now you can see we will mention also represented transform from camera view to the bird eye just you can see we will here implement this code okay in front of you here and mention the Al B ey points okay withth he EDC so now you can see we declared the all points and uh uh you will also see we are command one by one line and calibration points and coordinates points and also uh tracking points and create a directory of the vehical speed images okay you will also see it okay I mentioned this folder speed underscore images now you can see the starting to the cration frame FR per second from the zero and which frame per seconds will be captured one by one video you will see in also in this video okay now we will create the loop and each frame per second convert into all images into video and masking all video so I am quick overview of this algorithm and also we will put the bike track uh uh algorithm model Okay so you will see the all results in frame per seconds and mention in the frames okay so M show will be show the all the rectangles and you will see the display drawing the trail behind the vehicle and you also see the rectangle I will mention it when I'm run at this code uh so transform and drawing on curves okay so speed calculation will be have in uh this video will be xaxis y AIS coordinate and label to the speed calculation formula so also uh tracking generate for cars with the more than 120 km per hours so save the images into the your uh folder speedcore images okay so now write the all points in into mow and release to him show after video processing you will see this point is generet PDF format okay so draw images spending vehicle along with speed okay so what kind of we achieve in this algorithm we achieve the deduction part and you will get the one report okay so now we will add the logo also our transparent okay so here save all points this algorithm is uh in front of you will be happen Okay I save it and close it so I am requirement file is close it in front of you you will see the we are uh here this report this algorithm and uh calibration part and points mention the points and B ey View and uh you will see now how to run it step by step I'm close the terminal and you see the I am open it again in front of you and going to the CD desktop version directory sorry for the steam uh speed and uh uh open the project okay finally I am open the python app.py so I am showing to you this example this overview this whole story in front of you real time so there the mention the uh all tricks and uh here 1 2 3 4 okay in front of you object will be track and here above and below the 120 kilm hour speed and all formula will be happen in front of you it's distant formula and uh I am mention it one by one line and in front of you you will see the this demo okay after you can see uh here output will be released after you can see you will get the report okay so this report uh I think we need to delete it and he this will file will be right after the done our video so now we need to implement second demo video and uh I will mention it again this report okay here demo one number one I am just rename the video file path and close it contrl C and run it again python app.py so do not worry any type of question any type of you want to getting any you ask us so this happen in front of you you easily uh track the speed calculation uh it's very easy uh so now it's very simple app okay in front of you it's very simple app and you easily track the UR speed so I hope so now it's this report will be this example will be is simple for all of us and you also see the frame per seconds and here mention the name strike and also calibration the speed okay so guys I think it's I need to one one once again in front of you will be tested one more video from the download Google download we download from internet in front of you one video will be download and get the result and uh we will easily uh track this all points okay so you will also see the this type of video will be also dedi and speed calibration is very easy now this video is okay for us I think so we can download this video it's I think good for us or not uh let me Implement in our code and let's see and my internet speed is not good so let me create download this video in front of you just press the button and you will easily download it and after you download you can can see uh it's 1 minute will be released the video okay so after downloading we will get the result in our application so downloading is done and now you can see that this example uh I think it's not good but we will test it and I hope so it's working fine but uh uh I am not sure it so I'm just copy and uh going to the my desktop file okay in front of you I'm just past it and here name will be copy and paste and replace my previous video file okay so uh this step will be will be happen in front of you we will uh you will follow the steps and just run it and you can see we get the result and uh uh now you can see lot of zero here stop car it's zero and you will see the which type of speed will be uh going on so hope you like it and uh kindly support our channel for more getting informational updates every Saturday Sunday we will release these videos informational video so you also see uh our informational videos will be available uh our channel so hope you like it keep support us and uh see you in next video

Channel: Pyresearch

Views: 777

Keywords: Education, Pyresearch, Pyresearch - YouTube, how to Grow, money, facebook app, facebook, how to set goal, computervision, machinelearning, opencv, deeplearning, artificialintelligence, datascience, nlp, chatgpt4, ComputerVision, YOLOv8, ByteTrack, Supervision, ObjectTracking, SpeedEstimation, TechInnovation, RealTimeAnalysis, AdvancedTechnology, MachineLearning, DeepLearning, OpenSource, Innovation, TechAdvancement, DataAnalysis, SmartTech, DigitalTransformation, FutureTech, SoftwareDevelopment, Engineering, TechCommunity

Id: vdIOSJwgbEg

Length: 15min 7sec (907 seconds)

Published: Sun Apr 07 2024