Detection and Classification of Road Damage Based on Image Morphology and K-NN Method (K Nearest Neighbour)

Road pavement is a supporting factor for national development, especially in the distribution of trade goods and services as well movement human mobility. maintenance needs to be done regularly so that road always good condition, but weather loads are things cause damage. damage generally categorized into cracks, alligator cracks potholes. The purpose this research utilize image processing detect classify types steps involved include: acquisition with digital camera, conversion RGB images grayscale images, normalization, selection points, counting histogram bins, determining calculating noise morphology (closing opening) using disk element structure size 5, radial vector finally classifying K-NN (K Nearest Neighbor) method 3 classes K value 11. from classification results then calculated level based on category according SDI (Surface Distress Index) provisions, where crack seen width crack, percentage damaged area compared segment under review pathole many holes. test used 597 consisting 95% training data 5% data, obtained accuracy reached 83%.