Parallelizing Multiple Flow Accumulation Algorithm using CUDA and OpenACC.
Natalija StojanovicDragan StojanovicPublished in: ISPRS Int. J. Geo Inf. (2019)
Keyphrases
- learning algorithm
- experimental evaluation
- optimal solution
- worst case
- dynamic programming
- detection algorithm
- objective function
- recognition algorithm
- neural network
- search space
- preprocessing
- estimation algorithm
- improved algorithm
- knn
- memory requirements
- simulated annealing
- times faster
- matching algorithm
- computational complexity
- single pass
- path planning
- segmentation algorithm
- theoretical analysis
- computationally efficient
- linear programming
- high accuracy
- general purpose
- computational cost
- cost function
- k means