International Business Machines Corporation

Brahmaji Potu

Kevin Lynn Hill

The present invention provides a method and apparatus for transferring a video image, to be resized, from a host processor to an accelerator chip of a display adapter such that the storage capacity of a memory device in the chip is greatly reduced. The video data is first divided into M&times;M arrays of data elements. Then, the arrays are transferred one row at a time. Each row is stored before being processed by the chip. Consequently, since these rows are much shorter than the lines of frames of data elements, the storage capacity of the chip's memory device is greatly reduced.

Method and apparatus for block data transfer to reduce on-chip storage for interpolative video resizing

The present invention provides a method and apparatus for resizing a video image. In a preferred embodiment of the invention, an adapter having a bilinear filter is provided. The bilinear filter contains two first-in first-out (FIFO) memory devices for storing the video image and a vertical and horizontal filters for vertically and horizontally resizing the image. To resize the image, the video data representing the image is transferred to the graphics adapter in M.times.M arrays of pixel data. Each one of the (FIFO) memory devices has the capacity of storing one row of pixel data at a time. Adjacent rows are stored in the FIFO devices for vertical interpolation by the vertical filter. The oldest row in the FIFO devices is replaced by a new row for further interpolation. As each row is transferred to the horizontal filter, intermediate pixels are generated. In another embodiment, extra storage devices are used to hold data elements for interpolating between every two adjacent horizontal blocks.

Method and apparatus for resizing block ordered video image frames with

The present invention provides a method and apparatus for transferring a video image, to be resized, from a host processor to an accelerator chip of a display adapter such that the storage capacity of a memory device in the chip is greatly reduced. The video data is first divided into MxM arrays of data elements. Then, the arrays are transferred one row at a time. Each row is stored before being processed by the chip. Consequently, since these rows are much shorter than the lines of frames of data elements, the storage capacity of the chip's memory device is greatly reduced.

Method and apparatus for block data transfer to reduce on-chip storage

A video adapter for use in multimedia computer data processing system is disclosed. The video adapter is able to perform real time video resizing, such as video downscaling and upscaling, during compression and decompression playback within the multimedia computer data processing system. The data processing system typically includes a central processing unit, an input/output bus, and a memory controller for controlling a system memory and a level 2 cache for sending or storing data from the video adapter. The video adapter includes a bus enable logic, coupled to both the video decoder, video encoder, and field memory, that performs the video upscaling and downscaling. The bus enable logic also allows a video pass-through to allow compression and decompression of full resolution images without the need for upscaling or downscaling. The bus enable logic includes a clocked buffer, coupled to a U clock signal, a first enable gate, coupled in parallel to the clock buffer for passing either a U or V component of the digital video signal, and a second enable gate, coupled to a Y component of the digital video signal. The second enable gate is enabled by a Y enable signal and the bus enable log operates such that the Y component is active for two consecutive clock cycles and the U and V components alternate for the next two consecutive clock cycles such that in a given cycle period, two Y components are passed while one U and one V component are passed.

System for performing real-time video resizing in a data processing

A method and system in a data processing system are disclosed for improving video sequence image resolution when creating an enlarged final image from an initial image included within the video sequence. An enlargement factor is determined for creating a final image having a final image size from an initial image having an initial image size, where the enlargement factor is the quotient of the final image size divided by the initial image size. An intermediate factor is established to optimize the image resolution and minimizing the hardware required to produce the image. In response to the enlargement factor being less than the intermediate factor, the final image is created from the initial image utilizing the enlargement factor and bilinear interpolation. In response to the enlargement factor being greater than the intermediate factor, an intermediate image is created. The intermediate image has an intermediate size which is greater than the initial image size and less than the final image size. The intermediate image is created utilizing the initial image, the intermediate factor and bilinear interpolation. Thereafter, the final image is created utilizing the intermediate image, the enlargement factor, the intermediate factor, and nearest neighbor interpolation.

Growing community of inventors

Austin, TX, United States of America

Brahmaji Potu