Abstract: A hole-filling method, performable on a computing device, for providing a three-dimensional virtual model according to a technical aspect of the present application may comprises the operations of: acquiring an original training image and a hole creation training image, wherein the hole creation training image is an image in which at least one hole is created based on the original training image; creating a hole-filling training image by performing hole-filling on the hole creation training image using a neural network; performing spherical transformation on each of the hole-filling training image and the original training images; and training the neural network based on the difference between the spherically transformed hole-filling training image and the spherically transformed original training image.

Abstract: A computing device is provided. The computing device includes a memory configured to store one or more instructions; and a processor configured to execute the one or more instructions stored in the memory, in which the processor may execute the one or more instructions to generate a first predicted depth map image for a learning red/green/blue (RGB) image using a neural network, and train the neural network based on a difference between a learning depth map image having depth information and the first predicted depth map image. The learning depth map image may match 1:1 and may be generated based on the same setting change for a basic spherical image.

Abstract: According to a technical aspect of the present application, a texturing method, performable on a computing device, for creating a 3D virtual model based on a plurality of data sets, each of which is created from a plurality of shooting points in an indoor space and includes a color image and a depth image, may comprise: creating a 3D mesh model based on the created plurality of data sets; texturing each of the plurality of faces included in the 3D mesh model based on an association between the face and the color image; identifying a hole face that is displayed as a hole due to the absence of the association; and confirming a plurality of associated vertices associated with the hole face and setting a color of the hole face based on the colors of the confirmed plurality of associated vertices.

Abstract: A system for providing a virtual three-dimensional (3D) model according to a technical aspect of the present application includes a user terminal and a server. The user terminal derives relative movement information from a previous capturing point to each of a plurality of capturing points in a real indoor environment to generate location information about the corresponding capturing point and generates a 360-degree color image and a 360-degree depth map image on the basis of the corresponding capturing point to generate a capture dataset of the corresponding capturing point. The server receives a plurality of capture datasets each corresponding to the plurality of capturing points in the real indoor environment from the user terminal, relates a 360-degree color image to a 360-degree depth map image generated at each of the plurality of capturing points in accordance with locations of unit pixels, and sets a distance value and a color value per unit pixel to generate point groups.

Abstract: A hole-filling method, performable on a computing device, for providing a three-dimensional virtual model according to a technical aspect of the present application may comprises the operations of: acquiring an original training image and a hole creation training image, wherein the hole creation training image is an image in which at least one hole is created based on the original training image; creating a hole-filling training image by performing hole-filling on the hole creation training image using a neural network; performing spherical transformation on each of the hole-filling training image and the original training images; and training the neural network based on the difference between the spherically transformed hole-filling training image and the spherically transformed original training image.

Abstract: A texturing method of generating a three-dimensional (3D) virtual model is provided. The method is performed in a computing device, and includes acquiring an original learning image and a hole generation learning image for the original learning image, the hole generation learning image being an image in which at least one hole is generated based on the original learning image, generating a hole filling learning image by performing hole filling on the hole generation learning image using a neural network, performing spherical transformation on each of the hole filling learning image and the original learning image, and training the neural network based on a difference between the spherically transformed hole filling learning image and the spherically transformed original learning image.

Abstract: A computing device is provided. The computing device includes a memory configured to store one or more instructions; and a processor configured to execute the one or more instructions stored in the memory, in which the processor may execute the one or more instructions to generate a first predicted depth map image for a learning red/green/blue (RGB) image using a neural network, and train the neural network based on a difference between a learning depth map image having depth information and the first predicted depth map image. The learning depth map image may match 1:1 and may be generated based on the same setting change for a basic spherical image.