The use of object based classification with nDSM to increase the accuracy of building detection.
Baris BesolUgur AlganciElif SertelPublished in: SIU (2017)
Keyphrases
- classification accuracy
- accuracy rate
- detection accuracy
- detection rate
- support vector machine classifier
- robust detection
- false positives
- false negative
- machine learning
- predictive accuracy
- detection algorithm
- pattern recognition
- correct classification
- high accuracy
- feature extraction and classification
- pattern classification
- classification rate
- generalization ability
- error rate
- roc analysis
- support vector machine svm
- training set
- feature vectors
- classification systems
- classification scheme
- roc curve
- higher classification accuracy
- machine learning methods
- fold cross validation
- prediction accuracy
- detection method
- text classification
- feature extraction
- object detection
- support vector machine
- automatic detection
- computational complexity
- preprocessing
- individual classifiers
- confidence estimates
- neyman pearson
- confusion matrix
- digital mammograms
- feature selection
- training data
- discriminative classifiers
- computational cost
- supervised learning
- model selection
- automatic classification
- decision rules
- classification algorithm
- data sets