Progress for week 48 (2015)
From Robin
Contents |
Student template (copy this for your entry)
Budget
- Todo 1
- Todo 2
Accounting
- Done 1
- Done 2
Fredrik
Budget
- Full design in robdyn
- Start on Sferes2 integration?
Accounting
- Have full design with a good idea of how the system works
- Began working on a parameter-based environment (i.e generating terrain roughness and obstacles)
Eirik
Budget
- Implement functions for the image texture features entropy, standard deviation and range in C++.
- Optimize processing speed of the method for estimating road with single-gauss, based on the 5-dimensional feature space:
- Red-Channel
- Green-Channel
- Blue-Channel
- Illumination-Invariant Image
- Entropy
- (Standard Deviation)/(Range)
Accounting
- Done 1, skipped range-filter, because of simmilarity to standard deviation filter.
- Done 2, the algorithm can now process approximately 20 frames per second.
- Also discovered an error in my algorithm for forwarding samples, correction boosted accuracy somewhat.
Magnus
Budget
- Finish script for creating time series
- Finish script for performance measure
- Finish script for plotting results
Accounting
- Finished
- Started
- Started
- Also made sketch of the results i want
Per
Budget
- DAQ in SF
Accounting
- Done with most of the DAQ
- Due to some changes in the product I need some more data - My intern at ChemiSense will gather that for me - I have started a partnership with Morgam Kjølerbakken and Stian Aldrin at NILU and will probably do some work at their labs with supervision from Stian in machine learning in sensor analysis.
- Fitted regression for the temperature compensation and referance sensor for CO sensor.
- Write 3-5 pages about DAQ process