Abstract: A system for automatically placing virtual advertisements in sports videos, which includes a shot detection module, background extraction module, a calibration module and an asset placement module. The shot detection module detects the target shot of a sports video via a first trained model. The background extraction module performs a background extraction to extract the background of the first frame of the target shot to obtain a first background mask. The calibration module performs a camera calibration to detect a first transformation relation, between the first frame and a sport field template, via a second trained model. The asset placement module transforms an advertisement asset according to the first transformation relation to obtain a first transformed asset, and execute an asset placement to place the first transformed asset onto the first frame according to the first background mask to obtain a first image frame with placed advertisement.

Abstract: A processing device and method for a panoramic image includes converting the panoramic image to a cubemap format; under the cubemap format, calculating a depth information image of the panoramic image, and performing a seamless processing on the depth information image. The format of the depth information image is converted to the original format from the cubemap format to obtain the depth information image corresponding to the panoramic image.

Abstract: A system for automatically placing virtual advertisements in sports videos, which includes a shot detection module, background extraction module, a calibration module and an asset placement module. The shot detection module detects the target shot of a sports video via a first trained model. The background extraction module performs a background extraction to extract the background of the first frame of the target shot to obtain a first background mask. The calibration module performs a camera calibration to detect a first transformation relation, between the first frame and a sport field template, via a second trained model. The asset placement module transforms an advertisement asset according to the first transformation relation to obtain a first transformed asset, and execute an asset placement to place the first transformed asset onto the first frame according to the first background mask to obtain a first image frame with placed advertisement.

Abstract: A sports event video processing method is provided. The processing method includes: receiving a sports event input video; performing SOI detection on the sports event input video to obtain at least one SOI; performing logo detection and extraction on the at least one SOI to detect at least one logo; performing pixel-level rearrangement on the at least one detected logo; and outputting a sports event output video having completed pixel-level rearrangement.

Abstract: A cross-domain image comparison method and a cross-domain image comparison system are provided. The cross-domain image comparison method includes the following steps. Two videos in cross-domain are obtained. The videos are generated by different types of devices. A plurality of semantic segmentation areas are obtained from one frame of each of the videos. A region of interest pair (ROI pair) is obtained according to moving paths of the semantic segmentation areas in the videos. Two bounding boxes and two central points of the ROI pair are obtained. A similarity between the frames is obtained according to the bounding boxes and the central points.

Abstract: The present disclosure discloses a processing method and device for a panorama image. The processing method includes the following steps: A panorama image is converted a cubemap format. Under the cubemap format, a depth information image of the panorama image is calculated, and a seamless processing is performed on the depth information image. The format of the depth information image is converted to the original format from the cubemap format to obtain the depth information image corresponding to the panorama image.

Abstract: A cross-domain image comparison method and a cross-domain image comparison system are provided. The cross-domain image comparison method includes the following steps. Two videos in cross-domain are obtained. The videos are generated by different types of devices. A plurality of semantic segmentation areas are obtained from one frame of each of the videos. A region of interest pair (ROI pair) is obtained according to moving paths of the semantic segmentation areas in the videos. Two bounding boxes and two central points of the ROI pair are obtained. A similarity between the frames is obtained according to the bounding boxes and the central points.

Abstract: In a method for taking three-dimensional (3D) images, a camera unit of an electrical device is utilized to shoot a same scene with a same lens focal length and focusing on different distances respectively to generate several pictures. Depth-map information of several blocks on the scene is calculated according to levels of clarity on the blocks. Offsets of the blocks of the pictures are respectively adjusted for right and left eyes according to the depth-map information of the blocks to generate at least one 3D image. The present invention also discloses an electrical device for taking 3D images and a non-transitory computer-readable storage medium for storing the method for taking a 3D image.

Abstract: A computer-implemented method of creating a two dimensional code comprises providing a two dimensional code comprising a cell, providing a picture comprising a patch corresponding to the cell of the two dimensional code, providing a plurality of cell patterns different from each other, wherein each cell pattern comprises a plurality of sub-cells, and determining one of the plurality of cell patterns for the cell of the two dimensional code according to the patch of the picture.

Abstract: A computer-implemented method of creating a two dimensional code comprises providing a two dimensional code comprising a cell, providing a picture comprising a patch corresponding to the cell of the two dimensional code, providing a plurality of cell patterns different from each other, wherein each cell pattern comprises a plurality of sub-cells, and determining one of the plurality of cell patterns for the cell of the two dimensional code according to the patch of the picture.

Abstract: An image processing apparatus is for rendering a three-dimensional (3D) effect by transforming a first quadrilateral image to a second quadrilateral image. The apparatus includes a target image determining unit, a block determining unit and a graphic unit. The target image determining unit generates an outline associated with the second quadrilateral image according to the first quadrilateral image and the 3D effect. The block determining unit divides an area within the outline into a plurality of second blocks and correspondingly determines a plurality of first blocks from the first quadrilateral image. The graphic unit scales image data of the first blocks to respectively generate image data of the second blocks to obtain the second quadrilateral image.

Abstract: An image processing apparatus and an associated method capable of effectively reducing image artifact at an edge of an image without excessively compromising overall system performance. The method includes calculating an intersection point of an edge of the image and a pixel scan line, estimating a relative ratio of the image within a pixel block corresponding to the intersection point to generate an adjustment parameter, and adjusting a pixel value of the pixel block according to adjustment parameter.

Abstract: An image effect display method for displaying image effect during switching between a plurality of picture statuses includes providing a database, which includes a plurality of object models and a plurality of image effect templates; providing a status lookup table, which includes corresponding relationships between the plurality of picture statuses and the plurality of object models, and corresponding relationships between switching the picture statuses and the image effect templates; generating image effect data according to the database and the status lookup table by a transform program; and displaying a corresponding image effect according to the image effect data when one of the picture statuses is switched to another of the plurality of the picture statuses.

Abstract: An image processing apparatus and an associated method capable of effectively reducing image artifact at an edge of an image without excessively compromising overall system performance. The method includes calculating an intersection point of an edge of the image and a pixel scan line, estimating a relative ratio of the image within a pixel block corresponding to the intersection point to generate an adjustment parameter, and adjusting a pixel value of the pixel block according to adjustment parameter.

Abstract: An image processing method for rendering a three-dimensional by transforming a first quadrilateral image to a second quadrilateral image is provided, with at least one of the first and second quadrilateral images being a trapezoidal image. The image processing method includes providing the first quadrilateral image, generating coordinates of four vertices associated with the second quadrilateral image according to the first quadrilateral image and the three-dimensional effect; determining a relative height and a relative width according to a height and a width of the first quadrilateral image and the coordinates of the four vertices; and generating a plurality of pixel values of the second quadrilateral image with reference to a plurality of pixel values of the first quadrilateral image according to the relative height and the relative width.

Abstract: An image processing apparatus is for rendering a three-dimensional (3D) effect by transforming a first quadrilateral image to a second quadrilateral image. The apparatus includes a target image determining unit, a block determining unit and a graphic unit. The target image determining unit generates an outline associated with the second quadrilateral image according to the first quadrilateral image and the 3D effect. The block determining unit divides an area within the outline into a plurality of second blocks and correspondingly determines a plurality of first blocks from the first quadrilateral image. The graphic unit scales image data of the first blocks to respectively generate image data of the second blocks to obtain the second quadrilateral image.

Abstract: A method for image rendering is disclosed. A leftmost block of a top tile row (RTOP) is defined as an initial block (TINIT) and a reference pixel of the initial block is defined as an initial pixel. Block offset along the X and Y axes (?x and ?y) between a leftmost block of a current tile row and the initial block is calculated and attributes of a reference of the leftmost block is calculated using a formula ALEFT=AINIT+dA/dx*n*?x+dA/dy*n*?y. Block offset between a current block and the leftmost block (?z) is calculated, in which the current block and the leftmost block reside in the same tile row. Attributes of reference pixels of the current block are calculated using a formula ACURRENT=ALEFT+dA/dx*n*?z.

Abstract: A 3D-graphics processing method for processing 3D objects in a computer system defines a visible region having a far plane and a near plane. A clipping process is performed for a first object lying across the near plane while a second object lying across the far plane is exempted from the clipping process. In stead, the second object is performed with a rendering process as a whole to obtain a plurality of pixels. The depth values of the pixels are then compared with a depth value of the far plane. For any of the pixels having a depth value greater than the depth value of the far plane, it is discarded from display. On the other hand, any of the pixels of the first object not blocked by other pixels in front thereof and having a depth value smaller than the depth value of the far plane is outputted for display.

Abstract: A method and apparatus for generating shadow by using an adaptive shadow volume algorithm are disclosed. By accumulating the number of Z pass and Z fail during shadow volume rendering, the performance can be improved by adaptively selecting the z pass algorithm or the z fail algorithm. Accordingly, one of the algorithms (z-pass or z-fail) with better performance is used as the shadow volume algorithm for next frame.

Abstract: A core logic circuit which works with a CPU and a main graphics accelerator in a computer system is provided. The core logic chip includes a host controller electrically connected to the CPU for receiving a command from the CPU; an auxiliary graphing engine electrically connected to the host controller for receiving and processing the command; and a transmission controller electrically connected to the auxiliary graphing engine for transmitting the command that is processed and outputted by the auxiliary graphing engine to the main graphics accelerator to be further processed.