Joint Video Encoding for Dual Stream Stabilization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video processing methods require significant bandwidth and storage for both stabilized and non-stabilized video streams, especially in applications like body cameras, where both streams are needed for forensic purposes, leading to increased data communication and storage demands.

Innovation Solution

A video processing method that generates a joint encoded image stream by applying electronic image stabilization to one stream while bypassing the other, encoding intra frames based on the original stream, and encoding inter frames from both streams, allowing both stabilized and original streams to be available within a single joint stream, thereby reducing data requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If two separate video streams (stabilized and original) are encoded independently, then both streams are available for viewing, but bandwidth and storage requirements significantly increase

Engineering Contradiction:
Improveavailability of both stabilized and original video streamsVSAvoidbandwidth and storage requirements
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent merges the encoding of stabilized and original video streams into a single joint encoded bitstream. The encoder processes both streams simultaneously, sharing common components such as intra frames, motion estimation, and transformation processes. This combining approach maintains both video streams available for viewing while significantly reducing the total data volume compared to independent encoding, as common information is encoded once and reused for both streams

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The joint encoding system performs multiple functions within a single encoding process: it generates both stabilized and original video streams, applies motion compensation for both streams, performs transformation and quantization shared between streams, and produces a unified bitstream that contains both streams. This multi-functionality eliminates redundant processing and reduces overall bandwidth and storage requirements while ensuring both streams remain available

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Stability of the object's composition

If electronic image stabilization is applied to reduce camera shake, then image stability improves, but processing complexity and data requirements increase when both stabilized and original streams are needed

Engineering Contradiction:
Improveimage stabilityVSAvoidprocessing complexity for dual stream encoding
Core Design Contradiction:
Stability of the object's compositionVSDevice complexity

Solution Approach 1:

The patent combines the encoding processes for stabilized and original streams into a single unified encoder that handles both streams simultaneously. Common processing steps including intra frame encoding, motion estimation, motion compensation, transformation, and quantization are merged and shared between the two streams. This reduces processing complexity compared to running separate encoding processes, while still applying electronic image stabilization to achieve the required image stability

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The system performs preliminary stabilization processing on the video frames before the joint encoding stage. By pre-processing the frames to apply electronic image stabilization and determine motion vectors in advance, the actual encoding process can efficiently reuse this information for both stabilized and original streams, reducing the overall processing complexity during the main encoding operation

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11172124B2System and method for video processing
Publication Date: 2021.11.09 AXIS
  • US11172124B2 patent drawing
  • US11172124B2 patent drawing
  • US11172124B2 patent drawing

AI summary

A video processing comprises a video image processing pipeline and an encoder. The video image processing pipeline is processes images of captured video data and comprises: an image stream forming function which generates a first and second image streams each comprising a plurality of image frames, and an image stabilization function for image stabilizing the first image stream, thereby forming a stabilized image stream. The image stream forming function bypasses the second image stream from the image stabilization function, thereby forming an original image stream. The encoder encodes the stabilized image stream and the original image stream as a joint encoded image stream comprising encoded frames based on images frames from both the stabilized \the original image streams, wherein the joint encoded image stream comprises intra and inter frames, and wherein the intra frames are exclusively encoded based on image frames of the original image stream.