Network Surveillance Camera Metadata and Still Images for Video Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance camera systems struggle with inefficient and time-consuming video data searches due to the lack of real-time transmission of video analysis information and still images, necessitating separate decoding procedures.
Innovation Solution
A network surveillance camera system that transmits video analysis information as text-based metadata and still images in real-time using an extended RTP/RTSP streaming protocol, incorporating an RTP extension header with object IDs and coordinate information to facilitate quick and accurate video data retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video analysis information and still images are transmitted in real-time using extended RTP/RTSP protocol, then search efficiency and control convenience are enhanced, but device complexity and protocol complexity increase
Solution Approach 1:
The patent segments video data transmission by extracting and transmitting only relevant portions (still images of detected objects) alongside video analysis information, rather than transmitting entire video streams for search operations. This segmentation enables faster search efficiency while managing system complexity through selective data transmission.
Solution Approach 2:
The patent introduces an intermediary mechanism (the extended RTP/RTSP protocol with metadata and still image transmission) that mediates between the camera system and video receiving devices. This intermediary layer handles the complexity of real-time analysis information transmission, shielding the core video system while enabling enhanced search capabilities.
2Measurement precision
If separate video decoding procedures are performed for video data search, then search accuracy is maintained, but time consumption and processing overhead increase
Solution Approach 1:
The patent performs preliminary actions by generating and transmitting still images and video analysis information (metadata) in real-time alongside the video stream. These pre-processed elements contain key object information that can be searched without full video decoding, maintaining search accuracy while dramatically reducing the time required compared to decoding entire video sequences.
Solution Approach 2:
The patent creates simplified copies of video content in the form of still images and metadata that capture essential object information. These copies serve as searchable proxies for the full video data, maintaining sufficient accuracy for identification purposes while requiring minimal processing time compared to decoding complete video sequences.
Data Source
AI summary
Provided is a network surveillance camera system. The system comprises a camera for photographing a surveillance region to acquire video and audio signals for the surveillance region, and a video receiving device connected to the camera through network for receiving data from the camera in real time, wherein the camera comprises a metadata generation unit for generating video analysis information corresponding to the surveillance region as text-based metadata, and a still image generation unit for generating a still image by cropping a video portion corresponding to an identifiable object detected within the surveillance region from among the video analysis information.