Real-Time Search Method for Large-Scale Regional Targets Based on Parallel Google S2 Algorithm.
Fan WuQijie FengXiaoyong TangPublished in: HPCC/SmartCity/DSS (2019)
Keyphrases
- high accuracy
- computational cost
- search strategy
- detection method
- real time
- experimental evaluation
- cost function
- exhaustive search
- computational complexity
- improved algorithm
- k means
- theoretical analysis
- search procedure
- dynamic programming
- detection algorithm
- clustering method
- parallel implementation
- objective function
- optimization algorithm
- computationally efficient
- recognition algorithm
- significant improvement
- gpu accelerated
- probabilistic model
- input data
- estimation algorithm
- preprocessing
- optimization method
- segmentation algorithm
- similarity measure
- search space
- search mechanism
- synthetic and real images
- simulated annealing
- segmentation method
- convergence rate
- em algorithm
- hill climbing
- classification method
- depth first search
- learning algorithm
- matching algorithm
- support vector machine svm
- expectation maximization
- kalman filter
- tree structure
- search algorithm
- search methods
- energy function
- edge detection
- parallel version
- optimal solution
- pipelined architecture
- search results clustering
- alpha beta search
- color based tracking