CNN based semantic segmentation for urban traffic scenes using fisheye camera.
Liuyuan DengMing YangYeqiang QianChunxiang WangBing WangPublished in: Intelligent Vehicles Symposium (2017)
Keyphrases
- semantic segmentation
- traffic scenes
- ground plane
- field of view
- multiple cameras
- superpixels
- conditional random fields
- image sequences
- pedestrian detection
- object categories
- object detection
- camera calibration
- camera parameters
- multi camera
- video sequences
- light source
- vision system
- camera motion
- structure from motion
- multiple images
- object recognition
- high resolution
- multiple views
- multiple objects
- multi view
- image segmentation