Author: Alex Torres & co
Year: 2012, iSAIRAS conference
- Complex camera models that requires reevaluation of vision algorithm
- Fisheye lenses do not respect single viewpoint constraint (??)
- single viewpoint constraint (SVC) if each ray going through the effective pinhole would have crossed a single effective viewpoint
- Calibration with Scaramuzza method implemented on matlab vs Using a beam collimator
- Beam Collimator with gimbal generates a correction map for each pixel of the raw image according to a theoretical model, enables compensation for stereovision bench as well that leads to epipolar rectified images
- Omni Directional Vision Sensor HL-X52 for hypercatadioptric sensors
- Fish eye lens: Fujinon FE185CO57 that follows an equidistant projection
- Collimator calibration worked poorly with catadioptric but well with fisheye lens
- Figures of projection of the pixels on the ground thks to calibration parameters to compare both sensors
- Fisheye lenses were finally chosen however: lack of single effective viewpoint
- solid angle subtented by the pixel (??)
- Coaxial mount chosen -> column search for epipolar search (izi)
- Expansion factor (??)
Explication complète de l'aboutissement vers un montage coaxial stéréo fisheye, quelques points d'ombres à éclaircir avec alex