Motion Metadata Generation for Video Surveillance Search

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current video surveillance systems require users to pre-identify regions of interest before recording, which is impractical for unpredictable event locations, making it difficult to efficiently search for motion events in recorded video data.

Innovation Solution

The system processes video feeds in real-time by partitioning image frames into coarse-cells and macro-blocks, generating motion metadata that associates motion events with frame locations, allowing for efficient searching of motion events within specified regions post-recording.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If users pre-identify regions of interest before recording, then processing efficiency is improved, but adaptability to unpredictable event locations deteriorates

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidadaptability to unpredictable event locations
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The system performs preliminary action by partitioning the entire video frame into a grid of coarse-cells and further subdividing them into macro-blocks before recording. This pre-division structure is established in advance, allowing the system to efficiently locate and process motion events in any region without requiring users to pre-identify specific areas of interest. The preliminary grid structure enables both efficient processing and adaptability to unpredictable event locations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent applies segmentation by dividing the video frame into multiple coarse-cells, which are further divided into macro-blocks. This hierarchical segmentation creates a structured framework that allows the system to process motion detection at different levels of granularity. The segmented structure enables efficient processing by limiting analysis to relevant macro-blocks while maintaining adaptability to any event location within the segmented grid.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If motion metadata is generated for all video data, then search completeness is improved, but data processing overhead increases

Engineering Contradiction:
Improvesearch completenessVSAvoiddata processing overhead
Core Design Contradiction:
Loss of informationVSLoss of energy

Solution Approach 1:

The system segments the video data into coarse-cells and macro-blocks, generating motion metadata only for these segmented units rather than processing the entire video frame as a single unit. This segmentation reduces the volume of data requiring metadata generation while ensuring complete coverage of all potential motion events. The hierarchical segmentation allows for efficient processing overhead management.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts and generates motion metadata only for the macro-block level, which are the functional units containing motion information. By taking out and processing only the relevant macro-block data rather than all video pixels, the system reduces processing overhead while maintaining search completeness for motion events. The extraction of macro-block level metadata provides sufficient information for effective search without unnecessary processing.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS10515117B2Generating and reviewing motion metadata
Publication Date: 2019.12.24 CISCO TECHNOLOGY INC
  • US10515117B2 patent drawing
  • US10515117B2 patent drawing
  • US10515117B2 patent drawing

AI summary

Aspects of the instant disclosure relate to methods for generating motion metadata for a newly captured video feed. In some aspects, methods of the subject technology can include steps for recording a video feed using the video capture system, partitioning the image frames into a plurality of pixel blocks, and processing the image frames to detect one or more motion events. In some aspects, the method may further include steps for generating motion metadata describing each of the one or more motion events. Systems and computer-readable media are also provided.