This week our team finalized our decisions on parts we will be using for the project, and completed the order request forms. We mainly worked on preparing for design presentation that is coming up on next week. On the proposal presentation, we got feedback that our timing requirements were arbitrary. So, we worked with the professors and TAs to refine our timing requirements based on real chess game data. We also made sure that the board Demi will be designing meets the requirements for the CV processing (eg: it should have contrast between adjacent chess squares). We also made sure that we are budgeting our money correctly.
Yoorae’s Status Report for 10/02
I mainly spent this week preparing for the design presentation as a presenter that is coming up on next week. I started to set up an environment for OpenCV image processing, and started researching image detection algorithms we are going to use for the project. For the project, we are likely to use canny edge detection to detect the board squares and background subtraction algorithm to detect the player’s move (change in chess board pieces). I also started researching about code environment for Stockfish AI that we will be using for the project.
Yoorae’s Status Report for 9/25
For this week, our group met up and mainly discussed on finalizing parts decisions and making orders. I researched possible options for chess board and chess pieces to make an order. I spent my time mainly researching about which already available chess AI engine we are going to use for the project. I organized the possible options by complexity of a function (how easily we could understand it and apply changes), language it used, and the response time of an AI. We concluded that Stockfish or Sunfish will be the best possible options. I made an order for few parts. From now on, I will spend my time researching about CV and motion detection. As soon as Demi creates a sample chess board for testing and our camera arrives, we will begin testing.
Demi’s Status Report for 9/25
This week, I spent my time researching which LEDs to use. For our project, we only need 64 LEDs, one for each square of the 8×8 chessboard. There are two main requirements for our LEDs. The LEDs should be individually programmable to correctly display the AI’s move. Another requirement is that they should be able to transmit more than one color in order to deal with the ambiguity of castling, where four squares are lit up instead of two. We are going with WS2812B, which is the cheapest and most common LED strip that satisfies our needs.
Anoushka’s status report for 09/25
This week I worked on practising for my proposal presentation and preparing for questions. I also started looking more into the CV, and breaking it down into steps. The first step is edge detection and detecting the chessboard grid, which is what I am now focusing on. I also looked into cameras we could use for both Arduino and Rpi, and placed orders. I tried to find a chess board that would be similar enough to our chessboard to use while it is being built. I met with my teammates to discuss our different components and finalise them.
Team status report for 09/25
This week we worked on practising for the proposal presentation. After the proposal presentation, we began finalizing implementation details for our project. We thought about whether we should use Arduino or Rpi. We were going with RPi earlier but found that other teams had integration troubles with it. We met 3 times to figure out the details of each part of the project like LEDs, chessboard, CV, AI etc. The objective of the final meeting was to place order on equipment that we need. We also placed our order for a sample chessboard that we can get chess pieces from and also use to test our CV. We placed an order for an attachable logitech camera with 1280 x 720 resolution.
Yoorae’s Status Report for 9/18
This week, we mainly worked on as a team on refining our ideas for the project and preparing for the proposal presentation. We mainly discussed and worked on deciding which hard board pieces we are going to use, and organizing stuff that we need to order. When preparing for the presentation, we mainly discussed about specifying our quantitative data for testing metrics and refining on our testing methods. I personally researched on Chess AIs online that we might apply to our project.
Team status report for 09/18
This week we worked with course staff to finalise our project. We met with Prof. Yu and Funmbi on Monday and Wednesday to help narrow down the scope of our project. Earlier we had planned to develop the game playing AI ourselves, but it was decided that is too complex. So, we decided to use existing AI for the game.
We also got feedback on how to make sure that the project still remains unique and useful. This required us to make sure that the user actually gains something by using our board, and it’s not the same as using a regular Chess App.
We met three times during the week to iterate upon our ideas. We also worked on dividing up the project into chunks that could be handled independently. Then, we assigned tasks based off of our strengths. We also worked together on our proposal presentation slides.
Demi’s Status Report for 9/18
This week I worked with my group members to finalize our project idea. Based off Professor Yu’s suggestion, we now have two main areas: computer vision and creating a customized LED board. I will be working on designing and making the chessboard. I also worked on the proposal presentation slides and helped creating the Gantt chart.
Next week I will begin designing the chessboard. I hope to have a complete design by next week so I can start laser cutting the board the following week.
Anoushka’s status report for 09/18
This week I worked on finalising details of our project. I met with my teammates, Funmbi and Professor Yu. We discussed ways to narrow the scope of our project because our original plan covered too many areas. We decided to not create our own chess playing AI, and instead focus on Computer Vision, Signals and circuits. I will be working mainly on the Computer Vision side of the project, i.e. detecting the player’s moves on the chess board using timestamped images of it. I worked on dividing this task into sub-deliverables in the Gantt chart. I also began working on researching existing algorithms for chess piece detection and checking if they had anything in common with our use case. I didn’t find anything substantially similar, and am planning to begin implementing my own algorithm starting this week.
Additionally, I set up the WordPress site for the project. I also spent some time practicing for the proposal presentation on Monday.