• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.38 no.6
• /
• pp.635-644
• /
• 2020

Object recognition, detection and instance segmentation based on DL (Deep Learning) have being used in various practices, and mainly optical images are used as training data for DL models. The major objective of this paper is object segmentation and building detection by utilizing multimodal datasets as well as optical images for training Detectron2 model that is one of the improved R-CNN (Region-based Convolutional Neural Network). For the implementation, infrared aerial images, LiDAR data, and edges from the images, and Haralick features, that are representing statistical texture information, from LiDAR (Light Detection And Ranging) data were generated. The performance of the DL models depends on not only on the amount and characteristics of the training data, but also on the fusion method especially for the multimodal data. The results of segmenting objects and detecting buildings by applying hybrid fusion - which is a mixed method of early fusion and late fusion - results in a 32.65% improvement in building detection rate compared to training by optical image only. The experiments demonstrated complementary effect of the training multimodal data having unique characteristics and fusion strategy.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.2E
• /
• pp.5-11
• /
• 1995

In this paper, a method using a neural network was applied for the purpose of urilizing spatial features. The adopted model of neural network the three-layered architecture, and the training algorithm is the back-propagation algorithm. Co-occurrence matrix which is generated from original imge was used for imput pattern to the neural network in order to tolerate variations of patterns like rotation of displacement. Co-occurrence matrix is explained in appendix. To evaluate this method, classification was executed with this method and texture features method over the city area and sand area, which cannot be separated with the conventional method mentioned aboved. In the results of this method and texture features proposed by Haralick the method using texture features was separation rate of 67%~89%. On the contrary, the method using neural network proposed in this research was stable and high separation rate of 80%~98%.