Electronics and Telecommunications Research Institute

Jung-Jae Yu

Chang-Joon Park

Chang-Woo Chu

Hye-Sun Kim

Dong-Wan Ryoo

Kyung-Kyu Kang

Man-Hee Lee

Yun-Ji Ban

Jae-Hean Kim

Il-Kwon Jeong

Hye-Mi Kim

Ju-Won Lee

Disclosed herein are an apparatus and method for image outpainting based on learning of a single input image. The apparatus generates multiple extended images from an input image using multiple noise vectors, inputs a first extended image, among the multiple extended images, the input image, and an image acquired by horizontally flipping the input image and thereby trains a discriminator network such that the loss of a generative adversarial network is minimized, and learns the multiple extended images such that the sum of two differences is minimized, the first one being the difference between the input image and the center region of a second extended image, and the second one being the difference between images acquired by vertically segmenting the input image and flipping the segments and the left and right regions of a third extended image.

Apparatus and method for image outpainting based on learning of a single input image

Disclosed herein are a method and apparatus for matching 3D terrain information based on aerial images captured at different altitudes. The method includes receiving a high-altitude numerical height model based on a terrain image captured at a specific high altitude; receiving 3D terrain information observed from a low altitude, which is generated based on a terrain image captured at an altitude lower than the specific high altitude; generating a low-altitude numerical height model by converting the 3D terrain information into a numerical model in the same form as the high-altitude numerical height model; measuring the cross-correlation between the high-altitude numerical height model and the low-altitude numerical height model, thereby calculating matching parameters for enabling the low-altitude numerical height model to match the high-altitude numerical height model; and adjusting the geospatial coordinates of the 3D terrain information based on the matching parameters and outputting georeferenced 3D terrain information in the same coordinate system as the high-altitude numerical height model.

Method and apparatus for matching 3-dimensional terrain information using heterogeneous altitude aerial images

Disclosed herein are an apparatus and method for reconstructing a 3D model. The apparatus for reconstructing a 3D model includes an image acquisition unit for acquiring multi-view images by receiving image signals captured by multiple drones using cameras, a geometric calibration unit for estimating motion variables of the drones based on the acquired multi-view images, and a 3D model creation unit for reconstructing a 3D model of a dynamic object from the matched multi-view images using a multi-view stereo method.

Apparatus for reconstructing 3D model and method for using the same

A camera tracking apparatus for calculating in real time feature information and camera motion information based on an input image includes a global camera tracking unit for computing a global feature map having feature information on entire feature points; a local camera tracking unit for computing in real time a local feature map having feature information on a part of the entire feature points; a global feature map update unit for receiving the computed feature information from the global and local camera tracking units to update the global feature map; and a local feature selection unit for receiving the updated feature information from the global feature map update unit to select in real time the feature points contained in the local feature map. The local camera tracking unit computes the local feature map for each frame, while the global camera tracking unit computes the global feature map over frames.

Apparatus and method for real-time camera tracking

Growing community of inventors

Daejeon, South Korea

Jung-Jae Yu