Image processing device and frame buffer compressor
AssigneeSamsung Electronics Co., Ltd.
InventorSung Ho Jun
Life of this patent
Estimating the term…
Abstract
An image processing device includes a multimedia IP including a plurality of IP cores, at least one of which processes image data to generate source data, a frame buffer compressor, shared by the plurality of IP cores, compressing the source data to generate compressed data, and a memory coupled to the frame buffer compressor and accessed by the multimedia IP through the frame buffer compressor so that the compressed data is stored in the memory. The frame buffer compressor includes a quantization module which performs quantizing on the source data to generate recon data and a prediction module which executes intra-prediction on the recon data to generate prediction data.
Claims19
An image processing device comprising:
- a multimedia IP including a plurality of IP cores, at least one of which processes image data to generate source data;
- a frame buffer compressor shared by the plurality of IP cores and configured to compress the source data to generate compressed data;
- and
- a memory coupled to the frame buffer compressor and accessed by the multimedia IP through the frame buffer compressor, the memory configured to store the compressed data,
- wherein the frame buffer compressor includes a quantization module configured to perform quantizing on the source data to generate recon data, and a prediction module configured to execute intra-prediction on the recon data to generate prediction data, and
- wherein the source data comprises a plurality of source pixel data, and the quantization module is configured to perform a division operation on each of the plurality of source pixel data using a predefined, constant, quantization coefficient as a divisor to generate the recon data which correspond to quotients obtained by the division operation.
An image processing device comprising:
- a multimedia IP including a plurality of IP cores, at least one of which processes image data to generate source data;
- a frame buffer compressor shared by the plurality of IP cores and configured to compress the source data to generate compressed data;
- and
- a memory coupled to the frame buffer compressor and accessed by the multimedia IP through the frame buffer compressor, the memory configured to store the compressed data,
- wherein the frame buffer compressor includes a quantization module configured to perform quantizing on the source data to generate recon data, and a prediction module configured to execute intra-prediction on the recon data to generate prediction data, and
- wherein the source data comprises a plurality of source pixel data, and
- the quantization module is configured to perform a division operation on each of the plurality of source pixel data using a predefined, constant, quantization coefficient as a divisor to generate the recon data which correspond to quotients obtained by the division operation,
- wherein the frame buffer compressor further comprising:
- a mode selector which determines whether the frame buffer compressor operates in a lossy compression mode or a lossless compression mode,
- wherein, when the frame buffer compressor operates in the lossy compression mode, the quantization module generates the recon data from the source data, and the prediction module generates the prediction data from the recon data.
An image processing device comprising:
- a memory storing compressed data;
- a frame buffer compressor receiving source data and compressing the source data to generate the compressed data which are stored in the memory, and receiving the compressed data and decompressing the compressed data to generate output data;
- and
- a multimedia IP including a plurality of IP cores, the plurality of IP cores sharing the frame buffer compressor and at least one of the plurality of IP cores receiving the output data from the frame buffer compressor, accessing the memory through the frame buffer compressor and processing the output data to generate image data,
- wherein the frame buffer compressor is configured to compress the source data to generate the compressed data,
- wherein the frame buffer compressor includes a quantization module configured to perform quantizing on the source data to generate recon data, and a prediction module configured to execute intra-prediction on the recon data to generate prediction data, and
- wherein the frame buffer compressor is further configured to generate the prediction data based on the compressed data, and
- wherein the frame buffer compressor further includes a prediction compensation module configured to perform intra-prediction compensation on the prediction data to generate the recon data, and an inverse quantization module which performs inverse quantization on the recon data to generate the output data, and
- wherein the source data comprises a plurality of source pixel data, and
- the quantization module is configured to perform a division operation on each of the plurality of source pixel data using a predefined, constant, quantization coefficient as a divisor to generate the recon data which correspond to quotients obtained by the division operation.
Dependencies are inferred from the claim text.