Browse Prior Art Database

Method for efficient deringing using an additive pseudo-convolution mask

IP.com Disclosure Number: IPCOM000009892D
Publication Date: 2002-Sep-25
Document File: 4 page(s) / 178K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for efficient deringing using an additive pseudo-convolution mask. Benefits include improved performance and improved usability.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 50% of the total text.

Method for efficient deringing using an additive pseudo-convolution mask

Disclosed is a method for efficient deringing using an additive pseudo-convolution mask. Benefits include improved performance and improved usability.

Background

              Conventionally, a discrete cosine transform-based  (DCT-based) image/video data compression protocol, such as JPEG, MPEG-1/2/4 and H.261/263, provides excellent energy compaction properties. However, some unavoidable loss of high frequency components exists due to the quantization process. This loss results in ring-shaped noise patterns along with the edge component of the image and video signal.

              Deringing is one of the image and video post-processing methods developed for minimizing this ringing noise after decompressing image and video data (see Figure 1).

              Low pass filtering (LPF) plays a key role in a conventional deringing method. Combined with adaptive algorithms, LPF is widely accepted as an effective deringing method in many image/video compression standards. However, it has the following shortcomings:

•             High computational complexity

•             Lack of human visual characteristics

              Two-dimensional finite impulse response (FIR) filtering throughout the whole image/video frame requires a massive computation that can be increased with longer tap filters.

              Low pass filtering sometimes results in undesired blurring in output image/video, which can annoy a human viewer.

General description

      The disclosed method efficiently removes the ringing noise by using a simple additive pseudo-convolution mask in a DCT domain.

Advantages

              Some implementations of the disclosed structure and method provide one or more of the following advantages:

•             Improved performance due to reduced computational complexity

•             Improved performance and hardware friendliness due to not requiring multiplication is required

•             Improved performance and hardware friendliness due to the capability to be embedded into existing inverse discrete cosine transform (IDCT) modules

•             Improved usability due to being suitable for implement...