Structure from Motion help needed

asked 2014-02-18 09:07:44 -0500

pdf417 gravatar image

Hi All,

I have a sequence of images taken from a camera mounted on a mobile platform at a frame rate of 5fps and platform motion of around 2 to 5 centimeters per frame (the latter is just a speculation). The camera is looking at the ground with an angle of 41 degrees and therefore the horizon isn't visible.

I would like to build a structure from motion algorithm to help me in my analysis of the ground and objects (potential hazards) within the planned path. I've been chipping away at some Python code with OpenCV, but failed to get anywhere and I have now run out of options. I'm not a coder or programmer by any stretch but need to get this sorted ASAP in order to move onto the analysis (the core of my research).

I would therefore like to ask for someone with experience in Structure from Motion (SfM) using Python to get in touch and help me out against some sort of pre-agreed payment (I'm a student so please be kind with the remuneration you would be seeking). The outcome of this should be working python code that generates a crude Digital Elevation Model (DEM) from two subsequent images (not necessarily using every frame) and a detailed step by step explanation of what's going on in the code.

The idea isn't to buy code off you, but for me to learn about all the steps involved within the code. I am already familiar with the theory behind SfM but when it comes to converting that to code, I'm hopelessly stuck!

Please send me a private message on [email protected] with what you would expect to be paid and any other questions you might have.

This post might sound desperate, but only because I actually am!



edit retag flag offensive close merge delete