Content Adaptive De-Interlacing Algorithm一種內容自適應的去隔行算法
A new content adaptive de-interlacing (CADI) algorithm is designed to convert interlaced video to progressive video. Interlaced video has been widely adopted in current TV transmission standard. However, there are also several visual drawbacks like edge flicker, interline flicker, and line crawling. On the other hand, progressive video has been adopted by PC community and internet-processing tasks, since it does not have any visual drawbacks in interlaced video and simplifies many image-processing tasks. De-interlacing is needed to convert interlaced video to progressive video, for compatibility with existing TV transmission standard.
This invention, the CADI algorithm, works by interpolating the missing field based on the local image characteristics (smooth, horizontal edge, vertical edge and texture) and choose it most appropriate de-interlacing algorithm. A novel vector matching algorithm is also introduced for interpolating missing pixels in horizontal edge regions. As a result, it produces significantly high quality de-interlaced frame in terms of both subjective and objective quality in a wide range of video sequences. This CADI algorithm can adapt to local image characteristics and produces high quality de-interlaced frame in most situations. It can also adaptively change the search range according to the edge gradient and interpolate the pixels in the horizontal edge region accurately - functions not available in conventional algorithms.
1. High perceptual quality, especially for near-horizontal lines
2. Low complexity
- All progressive (non-interlace) display system
Patent Number: US7907209B2
Application Number: US2006433630A
Inventor: Au, Chi-Lim Oscar | Chong, Tak-Song
Priority Date: 13 May 2005
Priority Number: US7907209B2
Application Date: 12 May 2006
Publication Date: 15 Mar 2011
IPC Current: H04N000701
US Class: 348448 | 348452
Assignee Applicant: The Hong Kong University of Science & Technology
Title: Content adaptive de-interlacing algorithm
Usefulness: Content adaptive de-interlacing algorithm
Summary: Used for converting an image, with one row of pixels present and another row of pixels missing, to an image with two rows of pixels.
Novelty: Image converting method, involves selecting content adaptive de-interlacing algorithm based on result of determination, and computing measure of local horizontal and vertical high frequency activities for missing pixels
信息和通信技术/电信
图像处理
12 May 2006
US 11/433630
US 7907209
TTC.PA.254S
香港
