Video Image Pruning with Edge-Directed Interpolation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video encoding techniques face challenges in reducing bit-rate while maintaining high video quality, as methods like increasing quantization step size result in artifacts, and preprocessing techniques like data pruning can make decoded videos appear blurry due to loss of high-frequency information.

Innovation Solution

The proposed method involves pre-processing video images by selectively removing portions in the spatial domain, such as diagonal lines, prior to compression, and using edge-directed interpolation to restore the original image size after decoding, allowing for flexible bit-rate reduction without filtering pixels, thus preserving video quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If data pruning is performed to reduce bit-rate, then bit-rate is reduced, but video quality deteriorates due to loss of high frequency information

Engineering Contradiction:
Improvebit-rateVSAvoidvideo quality
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies preliminary action by performing data pruning before compression to reduce the amount of data that needs to be encoded. By selectively removing less important data (such as diagonal line portions) before the compression stage, the system reduces the overall bit-rate requirement while preserving the most critical visual information for quality maintenance.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements local quality by differentiating between different types of data in the video image. Instead of uniformly processing all data, the system identifies and treats different regions differently - specifically removing diagonal line portions while preserving other important data structures. This selective approach maintains local visual quality in regions that need it most while achieving overall bit-rate reduction.

Inventive Principle:
Principle #3Local quality

2Quantity of substance

If low-pass filters and down-sampling are used to reduce bit-rate, then bit-rate is reduced, but decoded video appears blurry

Engineering Contradiction:
Improvebit-rateVSAvoidimage sharpness
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies the extraction principle by selectively removing specific problematic data elements (diagonal line portions) from the image before compression. This targeted extraction differs from conventional low-pass filtering by removing only the specific data structures that cause blurriness while preserving the overall sharpness and detail of the decoded video.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of using conventional filtering approaches that blur the entire image to reduce bit-rate, the patent inverts the approach by selectively removing only the problematic diagonal line portions. This inversion of the conventional wisdom allows the system to reduce bit-rate without the side effect of image blurriness.

Inventive Principle:
Principle #13The other way round (Inversion)

3Area of stationary object

If conventional interpolation methods are used to restore image size, then image size is restored, but edges appear jerky

Engineering Contradiction:
Improveimage sizeVSAvoidedge smoothness
Core Design Contradiction:
Area of stationary objectVSManufacturing precision

Solution Approach 1:

The patent applies local quality by using edge-directed interpolation that adapts to the local characteristics of the image. Instead of applying a uniform interpolation method across the entire image, the system analyzes edge information and adjusts the interpolation process locally to match the actual image structure. This ensures smooth edges are preserved while maintaining the restored image size.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent implements feedback by using the decoded image information to guide the interpolation process. The edge-directed interpolation method incorporates feedback about the image's edge structure to adjust the reconstruction process, ensuring that interpolated regions blend smoothly with existing edges rather than creating jerky transitions.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2335419B1Methods and apparatus for video imaging pruning
Publication Date: 2020.11.04 INTERDIGITAL MADISON PATENT HLDG
  • EP2335419B1 patent drawingFigure 1~2
  • EP2335419B1 patent drawingFigure 3
  • EP2335419B1 patent drawingFigure 4

AI summary

Methods and apparatus are provided for video image pruning. An apparatus includes a data pruner (110) for pre-processing a picture prior to, and in preparation for, compression by encoding. The data pruner (110) selectively removes, in the spatial domain, at least one region within the picture. At the decoder end, an apparatus includes a a data restorer (140) for receiving a decompressed picture subsequent to decompression by decoding, and post-processing the decompressed picture by selectively restoring, in the spatial domain, at least one region in the decompressed picture based on information indicating a removal of the at least one region prior to a previously performed encoding process.