Leaning Impedance Distribution of Object from Images Using Fully Convolutional Neural Networks.
Masahiro KamigakiHisayoshi MuramatsuSeiichiro KatsuraPublished in: IECON (2020)
Keyphrases
- rigid body
- deformable objects
- image regions
- image analysis
- multiple objects
- image database
- three dimensional
- object recognition
- image data
- objects represented
- spatial distribution
- target object
- lighting conditions
- convolutional neural networks
- image registration
- normalized correlation
- image features
- visual appearance
- region of interest
- input image
- multiple images
- spatial relationships
- image matching
- image classification
- feature points
- edge detection
- keypoints
- image retrieval
- viewing angle
- acquired images
- geometric information
- convolutional network
- d objects
- illumination conditions
- partial occlusion
- image set
- test images
- surface shape
- multiscale
- segmentation algorithm
- segmentation method
- spatial information
- rigid transformation
- individual objects
- pixel level
- viewing direction
- object shape
- background clutter
- bounding box