diff --git a/src/computer_vision/README.md b/src/computer_vision/README.md
index 9e50966..1a9a0df 100644
--- a/src/computer_vision/README.md
+++ b/src/computer_vision/README.md
@@ -1,21 +1,57 @@
-# Setup
+# Computer Vision
+## Repository Organisation
+### computer_vision
+board_detection.py - Contains a ROS2 node that takes an image of a chessboard and identify the inner and outer corners of the chessboard, publishing whether coordinates have been found and their 2D image coordinates
+
+### config
+Contains camera configuration for the club webcam if using.
+
+### test
+Three basic tests and 3 images of random chessboards
+
+### train
+chess_piece_detection_training.py - Contains code to load data from roboflow and train a yolov8n model with it
+
+
+## Setup
 
 1. (Optional) If you are planning on training the YOLO model, copy `.roboflow_key.example` to `.roboflow_key`, get your Roboflow key and write it into the file
 2. Get camera calibration and parameter files and put them into `config/cam_params.yaml` and `config/camera_info.yaml`. This can be done in 2 ways:
 - Copy over the templates `config/cam_params_template.yaml` and `config/camera_info_template.yaml`. This is done by default in `install.py`.
-- Calibrate the camera by yourself (See the section below).
+- Calibrate the camera by yourself.
 
-## TODO: add documentation for how to calibrate the camera
-TODO: calibrate camera
-TODO: optimise params
 
-# Usage
-## Running the nodes
+## Usage
+### Running the nodes
 `board_detection` node doesn't use any extra parameters: just do `ros2 run computer_vision board_detection`
 
-## Debugging
+### Debugging
 To start the camera node directly, run `ros2 run usb_cam usb_cam_node_exe --ros-args --params-file config/cam_params.yaml`
 
-To simulate a camera with an image, run `ros2 run stub_scripts compressed_image_publisher --ros-args -p input_path:="test/test_images/chessboards/1.jpeg"`, where the image path can be replaced for other images
+To simulate a camera with an image, run `ros2 run stub_scripts compressed_image_publisher --ros-args -p input_path:="src/computer_vision/test/test_images/chessboards/1.jpeg", where the image path can be replaced for other images
 
 To view camera output, run `rqt`, select "Plugins/visualization/Image view" and select the correct topic
+
+
+## Resources
+### OpenCV
+OpenCV Tutorial - https://www.youtube.com/watch?v=oXlwWbU8l2o
+
+### YOLO
+How YOLO works - https://www.youtube.com/watch?v=svn9-xV7wjk
+YOLOv8 Model Specification - https://github.com/ultralytics/ultralytics/blob/main/docs/en/models/yolov8.md
+
+### Camera Calibration
+Monocular camera calibration using ROS - https://wiki.ros.org/camera_calibration/Tutorials/MonocularCalibration
+
+### Roboflow
+Roboflow Guide - https://blog.roboflow.com/train-yolov8-obb-model/
+
+### Ros Intergration
+YOLO integration - https://github.com/leggedrobotics/darknet_ros
+Ultralytics integration - https://docs.ultralytics.com/guides/ros-quickstart/#use-ultralytics-with-ros-sensor_msgsimage
+
+### Webcams
+Using Android phone as a webcam - https://adityatelange.in/blog/android-phone-webcam-linux/
+Physical Webcam model - https://amzn.eu/d/99ZbQrP
+
diff --git a/src/computer_vision/resource/computer_vision b/src/computer_vision/resource/computer_vision
deleted file mode 100644
index e69de29..0000000
diff --git a/src/computer_vision/train/chess_piece_detection.py b/src/computer_vision/train/chess_piece_detection_training.py
similarity index 89%
rename from src/computer_vision/train/chess_piece_detection.py
rename to src/computer_vision/train/chess_piece_detection_training.py
index 2449b7f..3fd1d66 100644
--- a/src/computer_vision/train/chess_piece_detection.py
+++ b/src/computer_vision/train/chess_piece_detection_training.py
@@ -4,7 +4,7 @@
 from ultralytics import YOLO
 
 try:
-    with open(".roboflow_key", "r") as f:
+    with open("src/computer_vision/config/.roboflow_key", "r") as f:
         ROBOFLOW_KEY = f.read().strip()
 except FileNotFoundError:
     from sys import exit, stderr
@@ -29,4 +29,4 @@
 
 model = YOLO("yolov8n.pt")
 
-results = model.train(data=f"{dataset.location}/data.yaml", epochs=100, imgsz=640)
+results = model.train(data=f"{dataset.location}/data.yaml", epochs=100, imgsz=640, device="cpu", batch=8)
\ No newline at end of file