- Oceanographic Data Processing
- Shape and Reflectance Acquisition from Scattering
- Smart Kitchen: What is in the refregerator?
- Super-Resolution Texture Mapping from Multiple View Images
- 4pi System
- An Accurate Shape Reconstruction
- Articulated Object Model Acquisition
The purpose of this research is to support human activities in fields that deeply relate to natural environments such as agriculture and fishery. We develop technologies for extracting valuable information from global-scale environmental data such as satellite images.
 Fishing Spot Estimation by Sea Temperature Pattern Learning, Masaaki Iiyama, Kei Zhao, Atsushi Hashimoto, Hidekazu Kasahara, Michihiko Minoh, Oceans'18,2018-05 slideshare
Most shape and reflectance acquisition methods use the light reflected from objectsf surfaces. When the reflection cannot be observed, e.g., when the objectfs surface is black matte or highly specular, its shape and reflectance are difficult to acquire. In this paper, we propose a method that can measure shape and reflectance with another approach. Our method involves the use of the scattering of reflected light in a participating media instead of using only the light reflection. We place the target in a participating medium and focus a laser beam on it. Cameras can observe the scattering of reflected light toward all reflection angles even if they can only observe light reflected toward the cameras.
We propose a method to monitor the grocery items in a home refrigerator. Using load sensors mounted under a shelf of the refrigerator, our method matches groceries that are put into and taken from the refrigerator. Our prototype load-sensing board uses four load sensors to acquire the weight and position of the grocery items, and we use this data to re-identify the groceries. Detailed experiments show that this feature can accurately re-identify grocery items and thus provide constant monitoring of refrigerator contents.
This paper presents an artifact-free superresolution texture mapping from multiple-view images. The multiple-view images are upscaled with a learning-based superresolution technique and are mapped onto a 3D mesh model. However, mapping multiple-view images onto a 3D model is not an easy task, because artifacts may appear when different upscaled images are mapped onto neighboring meshes. We define a cost function that becomes large when artifacts appear on neighboring meshes, and our method seeks the image-andmesh assignment that minimizes the cost function.
The 4pi measurement system is a system for reconstructing the volume of freely-moving objects. In our system, 36 cameras are arranged so that they can observe an object from all-round directions, while previous systems can only observe from upper viewpoints.
We are trying to accurately measure the 3D shape of an object using multiple cameras and lights.