Deteksi Crack Dan Studi Perbandingan Berdasarkan Faster R-CNN Dan Mask R-CNN

Nailah Rahdatul’Aisy (Universitas Pamulang)
Sella Amelia Putri (Universitas Pamulang)
Ranita Claudia Br Purba (Universitas Pamulang)
Muhammad Ikhwal (Universitas Pamulang)
Robbiul Jauza Erbila (Universitas Pamulang)
Perani Rosyani (Universitas Pamulang)

Biner : Jurnal Ilmu Komputer, Teknik dan Multimedia
Vol. 1 No. 3 (2023): BINER : Jurnal Ilmu Komputer, Teknik dan Multimedia

The intelligent gap detection method is an important guarantee for the realization of intelligent operation and maintenance, and it is very important for traffic safety. In recent years, the introduction of pavement cracks based on computer vision has attracted more and more attention. With the breakthrough of general deep learning algorithm technology in recent years, detection algorithms based on deep learning and convolutional neural networks have achieved better results in the field of crack recognition. In this paper, deep learning was investigated to intelligently detect road cracks, and Faster R-CNN and Mask R-CNN were compared and analyzed. Results show that the joint training strategy is very effective, and we can confirm that Faster R-CNN and Mask R-CNN complete the gap detection task when trained with only 130+ images and can outperform YOLOv3. However, joint strategy training led to a decrease in the effectiveness of the bounding box which was detected by the Mask R-CNN.