Industrial Technology Research Institute

Silicon Integrated Systems Corporation

Ruen-rone Lee

Chun-Kai Huang

Feng-Sheng Lin

Chung-Yen Lu

Wei-Kuo Chia

Li-shu Lu

Hsi-Jou Deng

Yu-Ming Huang

Ming-Fen Lin

Shih-Chin Lin

Yung-Feng Chiu

Ming-Mao Chiang

Kwo-Woei Yet

Yan-Jen Su

Sheng-han Wu

Feng-sheng Lin

Sheng-Han Wu

Shin-Ping Robert Wang

Kuo-Chang Fu

Chin Sung Lee

Chu-Fang Wang

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.

Method and system for sport game video processing

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.

System, non-transitory computer readable storage medium and method for automatically placing virtual advertisements in sports videos

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.

Image processing method and device for panoramic image

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.

Cross-domain image comparison method and system using semantic segmentation

An apparatus for controlling a stereo video display with non-stereo video source includes a memory, a read/write controller and a motion analyzer. The memory stores a current frame, a previous frame of the current frame, and a next frame for the current frame that is in preparation. The current frame and the previous frame are for a dextral image and a sinistral image of a stereoscopic image respectively. The read/write controller controls the addresses to write the current frame and the next frame into the memory, and the reading order of the current frame and the previous frame. The motion analyzer calculates an overall motion direction of the next frame in relation to the current frame, and compares the next overall motion direction to a previous overall motion direction of the current frame in relation to the previous frame, to control the action of the read/write controller to read from or write into the memory, so that when the horizontal relation between the next overall motion direction and the previous overall motion direction is changed, the corresponding relation between the current and the next frames and the dextral and the sinistral images is also changed.

Method and apparatus for controlling a stereo video display with non-stereo video source

The invention provides a method and an apparatus for dithering, and inversely dithering an image. The apparatus makes use of a pixel address and a pixel data to locate a dither reference value from a dither matrix and use the dither reference value to convert an original pixel data having N bits into a dithered pixel data having M bits (N&#x3e;M). On the other hand, the apparatus for performing an inversely dithering process makes use of the pixel address and the dithered pixel data (M bits) to locate a dither reference value from a dither matrix. Then, the dither reference value is employed to perform an inversely dithering process to recover the dithered pixel data into original pixel data.

Method and apparatus for dithering and inversely dithering in image processing and computer graphics

An apparatus for generating an interleaved stereo image includes a 3D graphics engine, an interleaved data merger and a register. The 3D graphics engine generates a shrunken left-eye image and a shrunken right-eye image of a stereo image based on video source data, the vertical sizes of the shrunken left-eye image and the shrunken right-eye image are shrunk compared with the vertical size of the video source data. The interleaved data merger stores the shrunken left-eye image and the shrunken right-eye image into two consecutive memory segments, and scans the memory segments while displaying the stereo image. A register controls a line number of a display region, so that the rows of the left-eye image shrunk and the right-eye image shrunk arrive at the display region in line-interleaved order. The invention also discloses a method for generating an interleaved stereo image.

Apparatus and method for generating an interleaved stereo image

A method for removing noise regions in a stereo 3D image, which includes a first eye image and a second eye image is achieved by calculating a maximum offset value and turning a horizontal synchronization signal and a display enable signal of the CRT timing parameters.

Method for removing noise regions in stereo 3D display system

An apparatus is provided to produce a real-time anaglyph, comprising a graphics engine, a memory and an anaglyph generator. The graphics engine provides a sync signal and generates a left eye image and a right eye image in accordance with a horizontal offset calculated from a 3D graphic animation, wherein the horizontal offset between the left and the right eye images creates illusion of depth. The left eye image and the right eye image are temporarily stored in the memory. The anaglyph generator then fetches the left eye image and the right eye image from the memory in response to the sync signal, for producing a filtered left image and a filtered right image. As such, each filtered image is respectively tinted with one of the complementary colors. By viewing through a pair of anaglyph glasses, the filtered left and the filtered right images can thus create perception of depth.

Apparatus for producing real-time anaglyphs

The invention provides a method of improving silhouette appearance in bump mapping, which not only reduces the operation overhead of applying displacement mapping to a whole model but also retains the truly geometric shape in displaying the object silhouette. The invention comprises the following steps: receiving, checking vertex, checking subdivision, subdividing, repeating, displacing and bump mapping. The invention also discloses a system employing the method of improving silhouette appearance in bump mapping.

Method and system of improving silhouette appearance in bump mapping

A system for full-scene anti-aliasing and stereo three-dimensional display control. The system includes a receiving buffer, a geometry offset generator, a rendering engine, and a blending engine. The receiving buffer receives a geometry, and the geometry offset generator receives stereo parameters and FSAA parameters, calculates first and second stereo offsets according to the stereo parameters, and calculates first and second FSAA offsets according to the FSAA parameters. The rendering engine combines the first and second stereo offsets with the first FSAA offset respectively to obtain first and second new offsets, combines the first and second stereo offsets with the second FSAA offset respectively to obtain third and fourth new offsets, and re-renders the geometry according to the first, second, third, and fourth new offsets respectively, so as to obtain first, second, third, and fourth new geometry. The blending engine blends the first and third new geometry to obtain a first display geometry, and blends the second and fourth new geometry to obtain a second display geometry.

System and method for full-scene anti-aliasing and stereo three-dimensional display control

The present invention discloses an apparatus and method for adjusting 3D transformation applied in a 3D computer graphic processing system, using a simple and practical computing method to simulate visual effect of human's eyes to separate input video into a left video and a right video. By the focus effect of the left video and the right video, a user can feel effect of changing depth of the input video, and creating 3D effect. Additionally, the present invention provides many parameters for convenience in adjusting the video depth and location in a display plane by users.

Apparatus and method for adjusting 3D stereo video transformation

A pixel based method for the computation of sub-pixel area-coverage is implemented in an area-coverage hardware module, within a 3D computer graphics rendering engine. Unlike the prior art segment based method which requires an operating aperture of an entire segment, the present invention only requires an operating aperture of one pixel. Therefore, the overall system pixel yield rate is increased.

Fast area-coverage computing method for anti-aliasing in graphics

A linear address organization for physically storing mip maps and rip maps in memory is disclosed. The subsampled data arrays of the mip maps and rip maps are sequentially stored in continuous subsequences of a continuous sequence of memory addresses. The subsequences of addresses are assigned in order of level of subsampling of the data arrays which make up the mip map or rip map. In the case of a mip map, the subsequences are assigned to the data arrays in order of increasing level of subsampling. In the case of rip maps, the data arrays are segregated into groups according to a first one of the two subsampling directions, such that each array in a particular group has the same level of subsampling in the first direction. Subsequences are assigned to each group of data arrays. The addresses of the subsequences are assigned within each group on a span by span basis where a span includes the texture data in a particular direction of each data array in order of increasing level in the second one of the two subsampling directions. Address generators are also disclosed for assisting in generating linear addresses to texture data with specific coordinates (u,v) in arrays having particular levels of subsampling.

Growing community of inventors

Hsinchu, Taiwan

Ruen-rone Lee