A Texture-Based Algorithm for Vehicle Area Segmentation Using the Support Vector Machine Method.
Ku-Jin KimSun-Mi ParkNakhoon BaekPublished in: RSFDGrC (2007)
Keyphrases
- segmentation algorithm
- segmentation method
- high accuracy
- support vector machine
- detection algorithm
- cost function
- detection method
- dynamic programming
- computational cost
- improved algorithm
- classification algorithm
- preprocessing
- significant improvement
- energy function
- objective function
- prior information
- clustering method
- k means
- support vector machine svm
- test images
- computational complexity
- final result
- optimization algorithm
- similarity measure
- image segmentation
- random walker
- tree structure
- watershed algorithm
- convergence rate
- gray level images
- matching algorithm
- classification method
- input data
- recognition algorithm
- segmentation result
- graph cuts
- reduced set
- gray value
- segmentation accuracy
- expectation maximization
- map matching
- learning algorithm
- vehicle license plate
- foreground background separation
- seed points
- probabilistic model
- simulated annealing
- multiple objects
- texture features