Introducing
Your new presentation assistant.
Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.
Trending searches
Matchday AI - Football
43
Matchday AI - Football
Features Implemented
Team 23
About Client :
Introduction
Matchday.ai is a sports tech company that provides high tech computer vision based solutions catered to broadcasters, pro athletes, semi professional and amateur athletes as well as sports teams/franchises. We use state of the art AI and Computer Vision techniques and our core technology is partly licensed from and validated by IIIT-H, which is a highly reputed university and ranked among the top 3 in Asia for AI research.
Team Members
Team Members :
The team consists of 4 :
-> Manas Kabre - 2018111014
-> Naman Baheti - 2018101099
-> Maheswara Rao - 2018101107
-> Yaswanth Balivada - 2018101126
TA Mentor - Prajwal Krishna Maitin
Client Name : Harshwardhan Kommana and Ganesh Y
Our Project
Our Project :
Football Match :
The match is where we get the data from
Data :
The data corresponding to the coordinates of the players and the ball
Visualization :
Different Visualizations (Dynamic and Static) of the game data we get
Problem Statement
Project
Introduction
The Task is to visualize a game of football given the database of the different coordinates of a player and the ball at every frame of the video. This is mainly so that players and coaches can analyze the game and make improvements accordingly. Also it targets the sport fans who love analyzing games.
Project Overview
Project Overview
The project mainly focuses on visualizing data related to football and badminton, as this helps to convert the visual tracked raw data into a more presentable form which helps coaches to analyze the data and make changes accordingly, as the data is no more raw and has been visualized. This product will help in bridging the gap between data and the coach. At an outcome level, raw data of a match must be presented in a format that gives the user all the functionality to visualize the game.
Project Description
Project Description
The projected is implemented in the form of a JavaScript based web app. It includes different visualization for a given game of football.
The Tools used are:
Static Visualization :
-> Heatmap.js to get heatmap visualization
-> Python - SimpleHTTPServer to host the app
-> Python for all scripting
-> HTML and CSS
-> JavaScript
Dynamic Visualization:
-> Python-Flask to host the app
-> Python Libraries (cv2, sqlite3,etc.)
Features Implemented :
Mainly there are two types of features :-
1) Static Visualization
2) Dynamic Visualization
So, for static visualization part, we are given a dataset corresponding to position of every player during a football match and we have to create a visualization of selected player.
For dynamic visualization part, we are given a match video and the dataset corresponding to player position and we have to create visualization on that video.
Dashboard :
Dashboard
-> Buttons of all functionality will be shown.
●->For Static part, data is fetched from the playerme.json file.
-> For Dynamic part, data is fetched from dummydata.json.
Heat Map
Heat map
-> User will be able to see heatmap corresponding to player he selected.
->User will be able to see heatmap of teams as a whole.
-> User will be able to change the timebar and heatmap will be changed correspondingly.
Point Map :
Point MAP
->User will be able to see the point at ball position for whole teams.
->User will be able to change the timebar and pointmap will be changed correspondingly.
Pass Map:
Pass Map
-> User will be able to see all passes connected by a line
-> User will be able to click on a particular pass and only the clicked pass will be shown.
->User will be able to change the time bar and passes will be changed
correspondingly.
Goals
Goals :
-> User can see all goals in the game.
-> A line will be shown connecting the player position with the goal.
Highlight Players
Highlight A player
-> User will be able to select the players whom he want to highlight.
-> User can select as many players as he wants.
-> A rectangular red colour box on position corresponding to selected player.
-> User will be able to dehighlight the player.
Highlight Formation
Highlight Formation
-> User will be able to select two or more than two players for highlighting.
-> A polygon connecting positions of all selected players will be shown.
-> User will be able to dehighlight the formation.
Curve formation
Curve formation
-> User will be able to select two points on screen.
-> User will be able to select the deviation in curve.
-> A curve corresponding to selected deviation(deviation => straight line) will be shown on screen.
UML Class Diagram :
Diagrams
UML Use Case DIAGRAM
UML USE CASE DIAGRAM :
UML State Diagram
UML State diagram
Client Feedback About Our Work
Client Feedback
Requirements Phase
-> We faced difficulty understanding project requirements.At that time the client gave us a clear idea of what he wanted multiple times.Like In start 2 meets he explained us requirements.
-> Client was also not clear about the requirements, so he used to change requirements in every meet. This caused a lot of problems.
Problems Faced
Implementation Phase
Implementation Phase
-> We did the project initially using python framework.But client rejected it claiming he wanted it running in a browser as an application.So we started coding again using html,css,javascript.One week of work got wasted here.
-> Since we didn’t know about various javascript libraries like d3.js, heatmap.js etc. , it took a lot of time for us to learn all things and do the needful.
-> During R2 we faced many issues in terms of communication with clients and our team due to Lock down in our places and sent away from college.
Crucial Learning
Crucial Learning
->Our project was related to data visualization. So, we had no knowledge of data visualization before. Hence, we got to know about data visualization and libraries related to data visualization in different programming languages i.e. python, javascript etc.
-> Since we were not familiar with flask and server related things, we learnt flask in this project.
-> This project was purely based on JavaScript except the flask thing, so we learnt many concepts related to JavaScript and many libraries i.e. d3.js, heatmap.js etc. of JavaScript.
-> Due to lockdown and problems caused due to COVID-19, there was a problem of communication b/w team members and the client also. But still we managed to meet and work together. So, we learnt about real life situations.