EvTemMap_basic branch

Python implementation for "Temporal-Mapping Photography for Event Cameras" -ECCV2024

Overview

The EvTemMap method deployed in this branch converts raw files collected by AT-DVS into TemMat, and performs the most basic time-domain to grayscale-domain conversion, generating grayscale images containing noise degradation.

Key Features

• Raw to TemMat Conversion: Transforms raw event data from AT-DVS into temporal matrices (TemMat).
• Time to Grayscale Conversion: Converts the temporal data into grayscale images, albeit with noise degradation.
• Adjustable Bias Parameter: The EvTemMap.py script includes a bias parameter which can be used to adjust the overall brightness of the grayscale images during visualization. Increasing the bias will result in brighter images.

Requirements

To run the EvTemMap method, ensure you have the following dependencies installed:

• numpy
• opencv-python
• metavision sdk (Please follow the installation guide at Prophesee Metavision Installation)

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/EvTemMap.git
   cd EvTemMap

2. Checkout to the EvTemMap_basic branch:

   git checkout EvTemMap_basic

3. Install the required dependencies:

   pip install numpy opencv-python

4. Follow the instructions in the Prophesee Metavision Installation Guide to install the Metavision SDK.

Usage

To run the EvTemMap basic method:

1. Open the EvTemMap.py file.
2. Adjust the bias parameter as needed to achieve the desired brightness for your grayscale images.
3. Execute the script (for normal brightness dataset):

   python EvTemMap.py --input path_to_raw_file

4. or Execute the script (for extremely low brightness dataset):

   python EvTemMap.py --input path_to_raw_file --bias 300e3