I want to test a visual odometry method for ground vehicles using a template matching algorithm. The method uses a downward facing camera perpendicular to the ground and estimates the motion of the vehicle by analyzing the image shift from frame to frame (30 fps at least).
Which kind of embedded camera do you recommend (low-cost) ?
What are the key points guiding the choice to reduce blur effect?
Artificial Retina is in the research development phase. It will solve in a natural way to such problems.
To reduce the blur effect you need to keep the exposure time short. That will give you darker images and you might consider external illumination. In that case, try to get the light distribution of the surface as even as possible. That is in fact more difficult than it sounds. Otherwise your algorithm will track the static light. Another option is to add filtering (such as LOG filter) to remove the effects of the static light.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Control Theory
- Control Systems