In the present invention, the following are provided: a 3D information generating unit (3) for generating 3D information as the data of a group of a plurality of points formed by projecting the values of respective pixels of a moving object in accordance with depth information detected from an image pickup image; an overlooking image generating unit (4) for generating an overlooking image by synthesizing the 3D information of the moving object with a space image of an image pickup target region; and a display control unit (5) for displaying the overlooking image. The present invention is configured so that, even in a case where there are multiple image pickup target regions in a large-scale building in which the floor configuration is complicated, it is unnecessary to display multiple image pickup images using split screen display, and by displaying one overlooking image formed by synthesizing the 3D information of the moving objects formed from the groups of a plurality of points with each of a plurality of image pickup target regions which are included in the entire space of the building, the overall state of the building can be ascertained in one glance by confirming the overlooking image.

The disclosed subject matter is directed to employing machine learning models configured to predict 3D data from 2D images using deep learning techniques to derive 3D data for the 2D images. In some embodiments, a method is provided that comprises employing, by a system comprising a processor, one or more three-dimensional data from two-dimensional data (3D-from-2D) neural network models to derive three-dimensional data from one or more two-dimensional images captured of an object or environment from a current perspective of the object or environment viewed on or through a display of the device. The method further comprises, determining, by the system, a position for integrating a graphical data object on or within a representation of the object or environment viewed on or through the display based on the current perspective and the three-dimensional data.

The disclosed subject matter is directed to employing machine learning models configured to predict 3D data from 2D images using deep learning techniques to derive 3D data for the 2D images. In some embodiments, a method is provided that comprises employing, by a system comprising a processor, one or more three-dimensional data from two-dimensional data (3D-from-2D) neural network models to derive three-dimensional data from one or more two-dimensional images captured of an object or environment from a current perspective of the object or environment viewed on or through a display of the device. The method further comprises, determining, by the system, a position for integrating a graphical data object on or within a representation of the object or environment viewed on or through the display based on the current perspective and the three-dimensional data.

Provided are a method and apparatus for decoding a multiview video and a method and apparatus for encoding a multiview video. The decoding method includes: receiving multiview image streams that configure the multiview video; obtaining, from a predetermined data unit header that includes information of a base-view picture included in a base-view image stream, picture order count (POC) information of the base-view picture; determining, using the POC information of the base-view picture, a POC of the base-view picture, based on an instantaneous decoding refresh (IDR) picture of a base-view; and determining, using the POC of the base-view picture, a POC of an additional-view picture that is included in a same access unit as the base-view picture and is transmitted.

An encoding device evaluates a plurality of processing and/or post-processing algorithms and/or methods to be applied to a video stream, and signals a selected method, algorithm, class or category of methods/algorithms either in an encoded bitstream or as side information related to the encoded bitstream. A decoding device or post-processor utilizes the signaled algorithm or selects an algorithm/method based on the signaled method or algorithm. The selection is based, for example, on availability of the algorithm/method at the decoder/post-processor and/or cost of implementation. The video stream may comprise, for example, downsampled multiplexed stereoscopic images and the selected algorithm may include any of upconversion and/or error correction techniques that contribute to a restoration of the downsampled images.

It is provided a method for detecting an object in a left view image and a right view image, comprising steps of receiving the left view image and the right view image; detecting a coarse region containing the object in one image of the left view image and the right view image; detecting the object within the detected coarse region in the one image; determining a coarse region in the other image of the left view image and the right view image based on the detected coarse region in the one image and offset relationship indicating position relationship of the object in a past left view image and a past right view image; and detecting the object within the determined coarse region in the other image.

The present disclosure relates to a method for transmitting two consecutive pairs of images. The method may include decimating each image with a ratio of 2, assembling the two decimated images of each pair in a composite image, transmitting the composite images, and reconstructing complete images from the composite images. In decimation, the information removed from the images of the first pair may be kept in the images of the second pair, from the spatial point of view, and the complete images may be reconstructed by de-interlacing processing from the composite images.

The invention discloses a display module with the divergence angle of an outgoing beam constrained again by the corresponding deflection aperture, which includes a multi-view display structure, a deflection-aperture array and a control device. The multi-view display structure includes a display screen, a light-splitting device, and a backlight-source assembly for providing backlights when a backlit-type display screen is adopted. The light-splitting device guides light beams from each group of pixels or sub-pixels to the corresponding viewing zone. A deflection aperture with a small size is designed for constraining the divergence angle of deflected light beams. Multiple deflection apertures play the function of enlarging the field of view, which is very limited when only a single deflection aperture exists. With orthogonal characteristics being assigned to deflection apertures for suppressing noise and/or for projecting more views, three-dimensional display with natural focus will get implemented with large field of view and low noises.

The invention discloses a display module with the divergence angle of an outgoing beam constrained again by the corresponding deflection aperture, which includes a multi-view display structure, a deflection-aperture array and a control device. The multi-view display structure includes a display screen, a light-splitting device, and a backlight-source assembly for providing backlights when a backlit-type display screen is adopted. The light-splitting device guides light beams from each group of pixels or sub-pixels to the corresponding viewing zone. A deflection aperture with a small size is designed for constraining the divergence angle of deflected light beams. Multiple deflection apertures play the function of enlarging the field of view, which is very limited when only a single deflection aperture exists. With orthogonal characteristics being assigned to deflection apertures for suppressing noise and/or for projecting more views, three-dimensional display with natural focus will get implemented with large field of view and low noises.

The disclosed subject matter is directed to employing machine learning models configured to predict 3D data from 2D images using deep learning techniques to derive 3D data for the 2D images. In some embodiments, a method is provided that comprises receiving, by a system comprising a processor, a panoramic image, and employing, by the system, a three-dimensional data from two-dimensional data (3D-from-2D) convolutional neural network model to derive three-dimensional data from the panoramic image, wherein the 3D-from-2D convolutional neural network model employs convolutional layers that wrap around the panoramic image as projected on a two-dimensional plane to facilitate deriving the three-dimensional data.