Time & Effort
Anthony Medina
Manuel Valdez
Week 13
7/25 - 7/29
Worked on Power point presentation
(10hrs)
Cleaned up website
(4hrs)
Testing and troubleshooting of raspberry pi
(6hr)
​
Week 12
7/18 - 7/22
Testing and troubleshooting of raspberry pi
(6hr)
Testing and troubleshooting of raspberry pi
(6hr)
​
Week 11
7/9 - 7/15
Finished first complete draft of project report
(12hrs)
​
Completed installation of project onto the raspberry pi
(12hr)
​
Week 10
7/4 - 7/8
Revised Chapter 2, drafted Chapters 3 and 4
(12hrs)
Troubleshooted PaddleOCR installation on raspberry pi
(12hr)
Researched converting our project into an android application
(8hr)
Week 9
6/27 - 7/1
Worked on issues installing the OCR module.
(6hrs)
Worked on chapter 2 of the report
(2hrs)
Researched alternate OCR modules
(2hrs)
Attempted installing PaddleOCR on the raspberry pi through multiple methods
(10hrs)
Researched alternate OCR modules and tested some of their results and processing speed.
(4hrs)
Week 8
6/20 - 6/24
Revised Chapter 1 and drafted Chapter 2
(4hrs)
Troubleshot the installation of the required modules on the raspberry pi
(6hrs)
Tested the OCR and object detection for accuracy and consistency.
(6hrs)
Troubleshot the installation of the required modules on the raspberry pi
(6hrs)
Week 7
6/13 - 6/17
Finished Chapter 1 of the report
(4hrs)
Setup and tested Text-To-Speech on the raspberry pi
2(hrs)
Installed PyTorch and YoloV5 to test new Object Detection Model.
8(hrs)
Implemented object localization model to separate objects for cropping and better OCR results.
(8hr)
Implemented algorithm to determine hand position compared to box location.
(4hr)
Week 6
6/6 - 6/10
Updated the Hardware Block diagram
(1hr)
Updated the Engineering Specifications
(1hr)
Constructed the hat
(2hrs)
Setup Bluetooth button and created a python script to control the camera with it
(6hrs)
Attempted to add general box detection to the model
(6hr)
Used OCR in conjunction with logo detection and product lists to estimate the brand and product
(6hr)
Week 5
5/30 - 6/3
Using labelimg I went through the first 7 products and labeled the logos on all of the images to prepare them for Tensorflow Model training
(4hrs)
Started assembling the Hat, using a 3D printer to create the housing to contain the Raspberry Pi, camera, and wiring.
(6hrs)
Tested various models using different types and combinations of Data augmentation.
(8hr)
Researched and created new model using PyTorch framework and YOLOv5.
(5hr)
Implemented OCR into our object detection.
(2hr)
Week 4
5/21 - 5/27
Created a python script to contain all essential functions and created a script to run object detection on images.
(4hrs)
Researched and setup MediaPipe for hand detection.
(4hr)
Trained new model with 14 logos.
(4hr)
Combined hand detection with logo detection.
(6hr)
Week 2
5/9 - 5/13
Assembled the raspberry pi, power supply, and camera. Installed the OS onto the raspberry pi and updated the software.
(4hr)
Fully setup a Tensorflow environment using Anaconda, and tested various model types for future training.
(6hr)
Week 3
5/16 - 5/20
Installed TensorFlow Lite and the Coco dataset, configured the pi camera and tested the object detection using the raspberry pi and the camera.
(6hr)
Trained first model to learn two logos, crest and Morton Salt.
(6hr)
Week 1
5/2 - 5/6
Ordered all required components as per our engineering specifications (1hr)
Began setting up Tensorflow and researching more information on creating a Tensorflow environment for PyCharm.
(3hr)