SPAD Regression Research

This repository contains various tests and jupyter notebook experimentations with various regression algorithms, mainly non-parametic, in order to better model the response curves of SPAD sensors from both EWH and EDH datasets

Motivation

The use of equi-depth histograms (EDH) to model photon arrival times in SPAD sensors for single-photon 3D cameras has greatly improved the accuracy of identifying the precise depth value of an object in a scene. Additionally, EDH also allow for much smaller binning requirements in order to accurately model depth levels, thus decreasing the amount of data needed to process each pixel in the image. Despite this, there are still issues with solely using EDH data to estimate depth and saturation, mainly the limited amount of actual depth levels that can be extracted from smaller-bin-sized datasets. The use of non-parametic regresssion equations as proposed both here and the writeup hope to remedy this by allowing for estimation of photon counts at any point in the camera's capture time. This not only allows for obtaining more information about the object captured by each pixel (such as reflectivity of the object and other sources of ambient light), but also allows for more potential depth levels for even smaller sized equi-depth bins.

Installation

The project can be installed using git via:

git clone https://github.com/elcosas/SPAD_Regression_Research.git

All libraries required for installation can be installed through pip using max-requirements.txt

pip install -r max-requirements.txt

or by using min-requirements.txt to avoid installing Jupyterlab dependencies

pip install -r min-requirements.txt

Usage (Notebooks)

First, once inside the project folder, use cd notebooks to change to the notebooks directory. The notebooks contain the various expermentations done with the sample data provided as well as different parametic and non-parametic regression methods tested with the sample data. the notebooks themselves can be viewed by running jupyter lab and then opening localhost:8888/lab/ in your prefered browser

Usage (Graph Tests)

The tests for the main three potential regression algorithms (GP, LOWESS, & KDE, as highlighted in the writeup) have been split into 2 folders for EDHs and EWHs, use cd {hist_type}_tests to move to the directory of your choosing
Each directory contains a file named hist_test_run.py which takes 3 arguments and can be used via the following syntax:

python3 hist_test_run.py {regression_method} {test_type} {data_set_#}

Where:
regression_method: the method to use, either "gaussian", "lowess", or "kde"
test_type: which test to run, either "single", "multi", or "all"
data_set_#: the data set to use, only accounted for with the single test but still required (default data set ranges from 0-29)
(TODO: provide example outputs)

Write-up installation

TODO

Attributions and License

TODO