Fast Transient Imaging
A Python framework revolving around reconstruction of impulse responses from AMCW lidar measurements.
Python-Framework for Fast Transient Imaging

This is an automatically generated documentation for the Python code that ships with the paper on "Solving Trigonometric Moment Problems for Fast Transient Imaging". Here are a few hints on where to start reading:

Installation

To use the framework you need Python 2.7 or newer (Python 3.x should be fine but has not been tested extensively) with NumPy, SciPy and matplotlib.

Some functionality of the framework has additional dependencies that do not need to be installed unless this functionality is used. If you want to store images as *.exr, you need OpenEXR bindings for Python. If you want to store images in various other formats (e.g. *.png), you need PIL.

Matlab

For users who prefer Matlab over Python we also provide Matlab implementations of the two most essential proposed algorithms. GetMaximumEntropySpectralEstimate.m is functionally equivalent to Reconstruction.GetMaximumEntropySpectralEstimate() and GetPisarenkoEstimate.m is functionally equivalent to Reconstruction.GetPisarenkoEstimate().

HLSL

The proposed algorithms map well to GPUs. To demonstrate this we provide an implementation of the maximum entropy spectral estimate in the High Level Shading Language (HLSL). It can be found in MaximumEntropySpectralEstimate.fx. Internally it uses Levinson's algorithm.

Data

AMCW lidar measurements in the form expected by Measurement.CMeasurement.AddCaptureData() can be found on the project webpage.