A computationally efficient 3D/2D registration method based on image gradient direction probability density function.
Soheil GhafurianIlker HacihalilogluDimitris N. MetaxasVirak TanKang LiPublished in: Neurocomputing (2017)
Keyphrases
- gradient direction
- probability density function
- density function
- segmentation method
- image registration
- density estimation
- edge detection
- grey level
- bayesian framework
- image matching
- similarity measure
- segmentation algorithm
- energy function
- em algorithm
- pixel values
- object recognition
- image features
- machine learning methods
- prior knowledge
- mutual information
- pixel intensities
- tensor field
- multiscale