the rants of a young software developer…
Posts tagged ‘Python’
I’ve been having trouble finding a good library or module for neural networking in Python. I am working on a project this semester: Facial Expression Recognition with the Future Goal of Detecting Micro-expressions.
If anyone has any advice, I would greatly appreciate it.
At first I was apprehensive when my boss asked me to start porting all of the lab’s software documentation to pdf and html using Sphinx (http://sphinx.pocoo.org/). It took a couple days to find a version that would install both on our local Widows machine as well as the Fedora server. I must say that following the rubric provided by Sphinx, the conversion was pretty straightforward.
Now that the majority of the information has been converted into reST (Restructured Text), I am looking forward to messing with the CSS for a customized look before publishing to the web.
Your comments or questions are much appreciated below!
So I was checking out SAGE (http://www.sagemath.org/) and it seems very cool. This guy certainly had a lot to say about why switching from MATLAB to SAGE was extremely beneficial. I agree with him for the most part. I would love to see all the software we are building to be converted into Python. Currently everything has been written in MatLab. Also now that we are ready to start publishing our work onto the web and collaborate with other genomic imaging labs throughout the country, we are trying to find the best way to do that. Python is free. The new GUI is going to be in Python. Everything else is in MatLab. This is going to be a lot of work. I’ve seen a couple different Python Libraries that will run MatLab syntax, but I believe our programs are a little too complex to run that way. We’ll see what the boss decides to do.
My research in the Image Processing Lab at Griffith University is going very well. I just recently completed a program that identifies blood vessels in retina scans. Over the next few weeks I will be testing the program in hopes to attain a high level of accuracy for vessel detection. I used the Python Image Library (http://bit.ly/1ikFGj) and NumPy (http://bit.ly/2P9kn). Using the Python programming language for biomedical research was something new for me because I before I had only used Python for game development.
Using the ideas of Elisa Ricci and Renzo Perfetti described in “Retinal Blood Vessel Segmentation Using Line Operators and Support Vector Classification” published in October, 2007 by IEEE Transactions on Medical Imaging, I believe my results are very close to the results they achieved.
While playing around with a car driving simulation I was creating for my game programming class, I devised a class that simulates the lateral frictional forces that act on a car at high speeds. Using the down arrow as a normal brake and the spacebar as a hand brake that “locks the wheels,” it is very easy to be thrown into a spin or skid out of control.
To do this, vectors must be used. The combination of velocity, momentum, acceleration, and friction all play a part on the direction the car is travelling. More importantly, the car does not always travel in the direction it is facing, so the sprite image of the car had to be drawn independant of the car’s velocity vector.