video camera security and surveillance system n.
Skip this Video
Loading SlideShow in 5 Seconds..
Video Camera Security and Surveillance System PowerPoint Presentation
Download Presentation
Video Camera Security and Surveillance System

Video Camera Security and Surveillance System

162 Vues Download Presentation
Télécharger la présentation

Video Camera Security and Surveillance System

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. Team Members : Semih Altınsoy - 04011501 Denis Kürov - 05011603 Team Advisor: Assist. Prof. M. ElifKarslıgil June, 2008 YILDIZ TECHNICAL UNIVERSITY COMPUTER ENGINEERING DEPARTMENT Video Camera Security and Surveillance System

  2. Introduction General System Modules Motion Detection Face Detection Face Recognition Network Conclusions and future work TABLE OF CONTENTS

  3. The system is a video surveillance security system which alerts users for some situations. The main workflow is that a fixed security camera will capture the videos continuously to checkout if there is movement in the area. If there is any movement in the zone, the movement differentiation between the current and previous one. If the diffrence is above the predefined value, the object will be checked if it is human or not. If the moving object is a human, its identity will be compared with current image database byface recognition module. If the face identity is not found in the database, the user will be alerted about current situation. Introduction

  4. That system’s abilities are; Burglar following Prevent accessing forbidden areas Child protection for accidents etc. introduction

  5. System has five main module. - Motion Detection - Face Detection Lighting Compensation Skin Tone Detection in YCbCr Space Find eyes and mouth Resize Face - Face Recognition ComputeEigenspace Project the Training Data Identify the Test Images - Network GENERAL SYSTEM MODULES

  6. General System Module

  7. Motion detection is a trigger for face detection and recognition module. For motion detection we are looking for current and previous frame. (Current Frame – Previous Frame) > Threshold For this situation, we can understand that there is a motion. Threshold is a value for motion sensitivity. Motion detection

  8. This technique is a face detection in YCbCr color space. Red-Green-Blue space is not a best choice for face detection. Firstly system corrects the color with a lighting compensation technique. It uses reference white to normalize the color appearance. The corrected RGB components are nonlinearly transformed in YCbCr color space. Face Detection based on the cluster (Cb/Y)-(Cr/Y) subspace. Skin-tone pixels are detected using an elliptical skin model in transformed space. Face detection

  9. C’i(Y)= { (Ci(Y) – Ci-(Y)) ∙ Wci / Wci(Y) + Ci-(Kh) if Y<Kl or Kh <Y { Ci(Y) if Y Є [Kl, Kh], Wci(Y)= { WLci + ((Y-Ymin) ∙ (Wci - WLci))/ (Kl - Ymin) if Y< Kl { WHci + ((Ymax-Y) ∙ (Wci - WHci))/ (Ymax- Kh) if Kh<Y Cb-(Y)= { 108 + ((Kl - Y) ∙ (118-108)) / (Kl - Ymin) if Y< Kl {108 + ((Y- Klh) ∙ (118-108)) / (Ymax- Kh) if Kh<Y Cr-(Y)= { 154 - ((Kl - Y) ∙ (154-144)) / (Kl - Ymin) if Y< Kl {154 - ((Y- Klh) ∙ (154-132)) / (Ymax- Kh) if Kh<Y We calculated transformed CbCr subspace with this transformation. Face detection

  10. And this is the ellipse formula for skin tones. x = (cosθ * C’b(Y)-cx) + (sinθ * C’r(Y)- cy) y= (-sinθ * C’b(Y)-cx) + (cosθ * C’r(Y)- cy) (x-ecx)2 / a2 + (y-ecy)2 / b2 = 1 Transformed CbCr Space Reference: Face Detection in Color Images - Rein-Lien Hsuy, Student Member, IEEE, Mohamed Abdel-Mottaleb, Member, IEEE, Anil K. Jainy, Fellow, IEEE Face detection

  11. Find Eyes and Mouth. The color of mouth region contains red component and weaker blue component. So the chrominance component Cr is stronger than Cb in mouth region. Mouth Map= Cr2 ∙ ( Cr2 – η ∙ Cr/Cb)2 η = 0.95 ∙ ((1/n) ∑Cr(x,y)2 ) / ( (1/n) ∑ Cr(x,y)/Cb(x,y) ) After the facial feature detection module rejects the regions that do not contain any facial features, we will find eyes of the face with black dots in the up-half of the picture. Reference: Face Detection in Color Images - Rein-Lien Hsuy, Student Member, IEEE, Mohamed Abdel-Mottaleb, Member, IEEE, Anil K. Jainy, Fellow, IEEE FACE DETECTION



  14. Eigenface (Training Stage)‏ Load The image data into memory Produce a centered image from all Create data matrix Create the covariance matrix Compute the eigenvalues and eigenspacesΩV = ΛV Order The EigenVectors : All the computed eigenvectors (except 0 valued ones) are sorted from high to low according to their eigenvalues. Each Training data is projected into eigenspace Every sub eigenspace stored in XML for future use Face Recognition

  15. Eigenface (Testing Stage)‏ Pull previous XML data into memory Test image is mean centredas we did in “Eigenspace Creation” section Mean centred image is projected into eigenspace The sub eigenspace is compared with all training ones If a very close match occurs the person is identified Otherwise you are a bad person,sorry! Face Recognition

  16. Picture in Database After that test we got 5 matches and 1misses which makes %83 success in that case. After that we repeated the same test with a person who is not in the database. We got only one match for the data set which makes the failure of the implementation to %17.

  17. Networking part (server) Collecting alert videos Provides safer Data transfer for clients Sends the videos and images to mobile and other clients Supplies log search and management for client parts No malicious users around ! Network

  18. Networking Part (Client and Server Case Study)‏ At the beginning Server produces its certificate(root)‏ Minions (clients) make server sign their certificates They are known clients and can send data to server A stranger is inside and detection part send alert with video Client passes the security of server and send its video Server now can alert the other minions that it signed before Everything is safer when it is crypted with 256 bit NETWORK

  19. In conclusion, this system can make burglar following, prevent accessing forbidden areas, child protection for accidents etc. This system is a real time system. So its some features like face recognition and detection can be better and more faster. Also for the future work there can be a mobile part. The system can be manageable by remote clients like mobile phones, PDAs and other PCs. In that way users can receive and control alerts of the system and manage logs, videos that system saved as critical. User can connect remote desktop computer with his/her mobile phone and control, manage logs and videos that system saved. Conclusions and future work

  20. THANK YOU ???