15.09.2011 Analysis, law enforcement, security, Uncategorized, video analysis, video analytics No Comments

Video Analysis/Analytics: Can we use it to detect criminal behaviors and activities?

I just found this report published by the National Criminal Justice Reference Service (NCJRS). Developed by Nils Krahnstoever, General Electric (GE) Global Research, it describes the development of a wide range of intelligent video capabilities relevant to law enforcement and corrections, and describes features of video surveillance that can help to enable early detection and possibly prevention of crimal incidents.

The study also points out, in a number of places, limitations of the technology, based on response activities and envronmental factors. it’s worth a read, here is the table of contents; you can read the document here Automated Detection and Prevention of Disorderly and Criminal Activities:

 Table of Contents

  • 1 Abstract
  • 2 Executive Summar
    • 2.1 Data Collection
    • 2.2 Crime Detection and Prevention
    • 2.3 System Evaluation and Feedback
    • 2.4 Law Enforcement Relevance and Impact
    • 2.5 Dissemination of Research Results
    • 2.6 Next Steps
  • 3 Introduction
  • 4 Data Sets and Data Collections 17
    • 4.1 GE Global Research Collection
    • 4.2 Airport and “Behave” Data
    • 4.3 Mock Prison Riot Data
      • 4.3.1 Venue
      • 4.3.2 Installation
      • 4.3.3 Camera Views
      • 4.3.4 Calibration
  • 5 Motion and Crowd Pattern Analysis 25
    • 5.1 Multi-camera Multi-target Tracking
    • 5.2 Detection and Tracking of Motion Groups
    • 5.3 Counting and Crowd Detection
    • 5.4 Simple Group-Level Events
    • 5.5 Group Interaction Model
    • 5.6 Group Formation and Dispersion
    • 5.7 Agitation and Fighting
    • 5.8 Advanced Aggression Detection
      • 5.8.1 Feature Tracking
      • 5.8.2 Motion Analysis
      • 5.8.3 Motion Classification and Clustering
      • 5.8.4 Results
  • 6 Identity Management
    • 6.1 PTZ Camera Control
      • 6.1.1 Introduction
      • 6.1.2 Related Work
      • 6.1.3 Experiments
      • 6.1.4 Discussions
    • 6.2 Identity Maintenance
  • 7 Social Network Estimation
    • 7.1 Introduction
    • 7.2 Experiments
    • 7.3 Conclusions
  • 8 Data Collection and System Testing at Mock Prison Riot 2009
    • 8.1 Collection and Testing Approach
    • 8.2 IRB Approval
    • 8.3 Collected Video Data
    • 8.4 Mock Prison Riot Detection and Tracking
    • 8.5 PTZ Control
    • 8.6 Behavior and Event Recognition
      • 8.6.1 Meeting / Approaching / Contraband Exchange
      • 8.6.2 Aggression Detection
      • 8.6.3 Fast Movement
      • 8.6.4 Distinct Group Detection
      • 8.6.5 Flanking Detection
      • 8.7 Performance Evaluation
      • 8.7.1 Sequence “Utah Leader Attack” (Nr. 00)
      • 8.7.2 Sequence “Utah Leader Attack 2” (Nr. 01)
      • 8.7.3 Sequence “Gang Killing other Gang” (Nr. 02)
      • 8.7.4 Sequence “Gang Killing other Gang 2” (Nr. 03)
      • 8.7.5 Sequence “Gang Killing other Gang 3 – Unrehearsed” (Nr. 04)
      • 8.7.6 Sequence “Aborted Attack” (Nr. 05)
      • 8.7.7 Sequence “Aborted Attack 2” (Nr. 06)
      • 8.7.8 Sequence “Gang Argument – Prisoners get attacked” (Nr. 07)
      • 8.7.9 Sequence “Gang Initiation” (Nr. 08)
      • 8.7.10 Sequence “Contraband Exchange” (Nr. 09)
      • 8.7.11 Sequence “Multiple Contraband Exchange” (Nr. 10)
      • 8.7.12 Sequence “Contraband with Fight” (Nr. 11)
      • 8.7.13 Sequence “Blended Transaction” (Nr. 12)
      • 8.7.14 Sequence “Shanking followed by Leaving” (Nr. 13)
      • 8.7.15 Sequence “Gang Hanging Out Followed By Several Fights” (Nr. 14)
      • 8.7.16 Sequence “Fight Followed by Guards Leading Offender Off” (Nr. 15)
      • 8.7.17 Sequence “Fight Followed by Guards Leading Offender Off” (Nr. 16)
      • 8.7.18 Sequence “Contraband – Officer Notices” (Nr. 17)
      • 8.7.19 Sequence “Argument Between Gangs – Officer Assault” (Nr. 18)
      • 8.7.20 Sequence “Contraband exchange followed by guard searching inmates” (Nr. 19)
      • 8.7.21 Sequence “Prisoner being attacked and guard intervening” (Nr. 20)
      • 8.7.22 Sequence “Fight breaking out between gang members and officers breaking it up” (Nr. 21)
      • 8.7.23 Sequence “Fight between gangs. Guards breaking fight up” (Nr. 22)
      • 8.7.24 Sequence “Fight between gangs. Guards breaking fight up” (Nr. 23)
      • 8.7.25 Sequence “Gangs fighting. Guards breaking fight up.” (Nr. 24)
  • A Public Dissemination
  • B Reviews and Meetings
    • B.1 Technical Working Group Meeting
    • B.2 Kick-Off Meeting at NIJ
    • B.3 Sensor and Surveillance Center of Excellence Visit
    • B.4 2008 Technologies for Critical Incident Preparedness Expo (TCIP)
    • B.5 Mock Prison Riot 2009
    • B.6 IEEE Conference on Computer Vision 2009
  • C Mock Prison Riot Data
    • C.1 Data Recorded while Processing
    • C.2 Sequences Processed in Detail
    • C.3 Data Recorded without Processing
  • D Techinical Details of the PTZ Camera Control
    • D.1 Problem Formulation
    • D.2 Objective Function
      • D.2.1 Quality Measures
      • D.2.2 Quality Objective
      • D.2.3 Temporal Quality Decay
    • D.3 Optimization
      • D.3.1 Asynchronous Optimization
      • D.3.2 Combinatorial Search
  • E Techinical Details of Social Network Analysis 110
    • E.1 Building Social Network
      • E.1.1 Face-to-Track Association via Graph-Cut
    • E.2 Discovering Community Structure via Modularity-Cut
      • E.2.1 Dividing into Two Social Groups
      • E.2.2 Dividing into Multiple Social Groups
    • E.2.3 Eigen-Leaders

 

Comments are closed.