Wearable Camera Video Analysis for Suspect Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for police officers using wearable cameras to identify suspects in recorded videos are inefficient, requiring manual visual inspection of multiple video files, which is time-consuming and impractical for quick suspect identification.
Innovation Solution
A monitoring video analysis system and method that utilizes a wearable camera and a server to generate and record videos in predetermined time segments, allowing the server to analyze and display a list of objects, such as faces, appearing across multiple video files, facilitating quick identification of suspects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual visual inspection of video frames is used to identify suspects, then thorough examination of all video data is possible, but the search time becomes excessively long and productivity decreases
Solution Approach 1:
The patent replaces the mechanical manual visual inspection process with an automated computer-based image processing system. The server automatically extracts frames from multiple video files, detects faces using image processing algorithms, and identifies suspects appearing across multiple files, thereby eliminating the time-consuming manual review process while maintaining identification accuracy
Solution Approach 2:
The patent introduces an intermediary automated analysis system between the raw video data and the final suspect identification. This intermediary system performs frame extraction, face detection, and cross-video correlation analysis, serving as a bridge that processes the large volume of video data efficiently before presenting results to investigators
2Reliability
If all video data files are processed to ensure complete suspect identification, then identification reliability improves, but the time and computational resources required increase significantly
Solution Approach 1:
The patent segments the video analysis process into distinct automated stages: frame extraction from multiple video files, face detection in each frame, face matching across frames and files, and suspect identification. This segmentation allows systematic processing of all video data to ensure reliability while optimizing each stage for efficiency
Solution Approach 2:
The patent performs preliminary automated processing of video frames before human review, including automatic frame extraction, face detection, and pre-screening for individuals appearing in multiple files. This preliminary action filters out false positives and prepares organized candidate lists, ensuring thorough examination while reducing the time investigators need to spend on initial screening
Data Source
AI summary
A monitoring video analysis system includes a wearable camera that is held or possessed by a user, and a back end server that receives a captured video. The wearable camera records a captured video by capturing a video of a scene for each ten minutes. The back end server determines a face of a person appearing in n (n is an integer of two or more) captured videos by using a plurality of captured videos of the scene for each 10 minutes, and displays a thumbnail list UI in which the determined face of the person is registered, on a monitor.