Geoinformation Science Research Team

3DDB Viewer Launched

3DDB Viewer is a web user interface for searching, visualizing, and downloading various 3D data (e.g. point clouds, structures, mesh etc.).

3DDB Viewer
Click here for the manual.

MUltiband Satellite Imagery for object Classification (MUSIC) for HotArea dataset

We created MUltiband Satellite Imagery for object Classification (MUSIC) for HotArea (HA) dataset. The HA is a system to automatically detect hotspots (e.g., fires and volcanoes) in the world in mid-resolution satellite data and displays the results on a web-based GIS system. Currently Hotarea utilizes Landsat 8 and Sentinel-2 data in global scale and in some selected regions, respectively.

MUltiband Satellite Imagery for object Classification (MUSIC) to detect Golf Course

We have created MUltiband Satellite Imagery for object Classification (MUSIC) to detect Golf Course (GC). There are more than 30,000 golf courses all over the world. They are good targets for research of automatic object detection due to their specific shape, world-wide distribution and common size. MUSIC for GC is the dataset to support the global survey of golf courses based on satellite imagery.

Transfer Learning With CNNs for Segmentation of PALSAR-2 Power Decomposition Components

Poliyapram Vinayaraj , Ryu Sugimoto, Ryosuke Nakamura and Yoshio Yamaguchi

<Abstract>
Water/ice/land region segmentation is an important task for remote sensing, as it analyses the occurrence of water or ice on the earth's surface. Many previous deep learning researches effectively utilized multispectral satellite images for highly accurate water/ice/land region segmentation. However, the deep-learning-based segmentation of synthetic aperture radar images still remains a challenging task due to the unavailability of enough labeled data. In order to overcome this issue, we designed a two-step deep-learning-based transfer learning model that needs a very limited number of labeled samples. The proposed approach consists of two models. The first model is a deep encoder-decoder 6SD to Landsat-8 multispectral translation model (DTF) that translates fully polarimetric PALSAR-2 6SD data to six new features. As for the second model (transfer learning), it utilizes the DTF features to fine-tune the model using the Landsat-8 multispectral pretrained model for water/ice/land segmentation. Hereinafter, the proposed two-step model is referred to as DTF-TL. Also, a qualitative and quantitative analysis was carried out to evaluate the performance of the proposed model (DTF-TL) and compare it with various transfer learning methods. Overall, the DTF-TL model outperformed the other models with consistent and reliable water/ice/land segmentation results in terms of the recall (0.980), precision (0.981), F1-score (0.981), mean intersection over union (0.962), and accuracy (0.989).

Transfer Learning With CNNs for Segmentation of PALSAR-2 Power Decomposition Components

Poliyapram Vinayaraj , Ryu Sugimoto, Ryosuke Nakamura and Yoshio Yamaguchi

<Abstract>
Water/ice/land region segmentation is an important task for remote sensing, as it analyses the occurrence of water or ice on the earth's surface. Many previous deep learning researches effectively utilized multispectral satellite images for highly accurate water/ice/land region segmentation. However, the deep-learning-based segmentation of synthetic aperture radar images still remains a challenging task due to the unavailability of enough labeled data. In order to overcome this issue, we designed a two-step deep-learning-based transfer learning model that needs a very limited number of labeled samples. The proposed approach consists of two models. The first model is a deep encoder-decoder 6SD to Landsat-8 multispectral translation model (DTF) that translates fully polarimetric PALSAR-2 6SD data to six new features. As for the second model (transfer learning), it utilizes the DTF features to fine-tune the model using the Landsat-8 multispectral pretrained model for water/ice/land segmentation. Hereinafter, the proposed two-step model is referred to as DTF-TL. Also, a qualitative and quantitative analysis was carried out to evaluate the performance of the proposed model (DTF-TL) and compare it with various transfer learning methods. Overall, the DTF-TL model outperformed the other models with consistent and reliable water/ice/land segmentation results in terms of the recall (0.980), precision (0.981), F1-score (0.981), mean intersection over union (0.962), and accuracy (0.989).

Canopy Averaged Chlorophyll Content Prediction of Pear Trees using Convolutional Auto-Encoder on Hyperspectral Data

Subir Paul, Vinayaraj Poliyapram, Nevrez İmamoğlu, Kuniaki Uto, Ryosuke Nakamura, D. Nagesh Kumar

<Abstract>
Chlorophyll content is one of the essential parameters to assess the growth process of the fruit trees. This present study developed a model for estimation of canopy averaged chlorophyll content (CACC) of pear trees using the convolutional auto-encoder (CAE) features of hyperspectral data. This study also demonstrated the inspection of anomaly among the trees by employing multi-dimensional scaling (MDS) on the CAE features and detected outlier trees prior to fit nonlinear regression models. These outlier trees were excluded from the further experiments which helped in improving the prediction performance of CACC. Gaussian process regression (GPR) and support vector regression (SVR) techniques were investigated as nonlinear regression models and used for prediction of CACC. The CAE features were proven to be providing better prediction of CACC when compared with the direct use of hyperspectral bands or vegetation indices as predictors. The CACC prediction performance was improved with the exclusion of the outlier trees during training of the regression models. It was evident from the experiments that GPR could predict the CACC with better accuracy compared to SVR. In addition, the reliability of the tree canopy masks, which were utilized for averaging the features' values for a particular tree, was also evaluated.