Project Login
Registration No:

Coupled Kernel Embedding for Low-Resolution Face Image Recognition

Platform : DOT NET

IEEE Projects Years : 2012 - 13

Coupled Kernel Embedding for Low-Resolution

Face Image Recognition




Practical video scene and face recognition systems are sometimes confronted with low-resolution (LR) images. The faces may be very small even if the video is clear, thus it is difficult to directly measure the similarity between the faces and the high-resolution (HR) training samples. Face recognition based on traditional super-resolution (SR) methods usually have limited performance because the target of SR may not be consistent with that of classification, and time-consuming SR algorithms are not suitable for real-time applications. In this paper, a new feature extraction method called coupled kernel embedding (CKE) is proposed for LR face recognition without any SR preprocessing. In this method, the final kernel matrix is constructed by concatenating two individual kernel matrices in the diagonal direction, and the (semi)positively definite properties are preserved for optimization. CKE addresses the problem of comparing multimodal data that are difficult for conventional methods in practice due to the lack of an efficient similarity measure. Particularly, different kernel types (e.g., linear, Gaussian, polynomial) can be integrated into a uniform optimization objective, which cannot be achieved by simple linear methods. CKE solves this problem by minimizing the dissimilarities captured by their kernel Gram matrices in the LR and HR spaces. In the implementation, the nonlinear objective function is minimized by a generalized eigenvalue decomposition. Experiments on benchmark and real databases show that our CKE method indeed improves the recognition performance.






We design and develop a low-cost based human-tracking system. Face Recognition (FR) systems are increasingly gaining more importance. Face detection and tracking in a complex scene forms the first step in building a practical FR system. We detect and track human faces in color image sequences. Skin color classification and morphological segmentation is used to detect face(s) in the first frame. These detected faces are tracked over subsequent frames by using the position of the faces in the first frame as the marker and detecting for skin in the localized region. Specific advantages of this approach are that skin color analysis method is simple and powerful, and the system can be used to detect/track multiple faces.




This work presents a real time system for detection and tracking of facial features in video sequences. Such system may be used in bank and in places where security is is more important. We have used a statistical skin-color model to segment face-candidate regions in the image. The presence or absence of a face in each region is verified by means of an eye detector, based on an efficient template matching scheme . Once a face is detected, the pupils, nostrils and lip corners are located and these facial features are tracked in the image sequence, performing real time processing.




































Human motion tracking is primarily concerned with determining the existence and location of humans within certain regions of space. Automatic face recognition is a process of identifying a test face image with one of the faces stored in a prepared face database. Real world images need not necessarily contain isolated face(s) that can directly serve as inputs to a FR system. Hence, there is a need to isolate or segment facial regions to be fed to a FR system. Most of the time, a video sequence of the scene is available using which a person may have to be recognized.




 For recognition, we need the face position in which it is best recognizable by the present day FR algorithms. Hence, a robust system that detects and tracks a face is necessary. Face detection and tracking becomes an important task with the growing demand for content-based image functionality. Though human beings detect/track faces with very little effort, it is not easy to train a computer to do so. In pattern recognition parlance, human face is a complex pattern.




Different poses and gestures of the face accentuate complexity. The detection scheme must operate flexibly and reliably regardless of the lighting conditions, background clutter in the image, multiple faces in the image, as well as variations in face scale, pose and expression. The system should be able to detect the face even under small occlusions.




Over the past several decades, surveillance techniques have matured dramatically. Analog tapes and security personnel are being replaced with Internet Protocol (IP) technology, leveraging digital video cameras, remote access, and intelligent analytics. This evolution provides organizations with significant opportunities to improve security and reduce operating costs.


Today’s businesses and public agencies are faced with a critical need to protect employees, clients, citizens and assets from possible threats with a security system that enables rapid response to security breaches and prompt investigation of events. Organizations are additionally challenged with managing tremendous amounts of information in various forms, including video, voice, electronic data and paper.




The goal of this technological and commercial intelligence report is to describe the intelligent video surveillance sector for the security of individuals and places. It is an emerging, little known technology that is changing how traditional video surveillance is used and is opening up a world of opportunities, making it possible to foresee new market segments emerging in the security sector.




This document, intended for a non-expert audience, discusses the ins and outs of this technology and tries to characterize the market it represents, not only globally, but more specifically in Quebec. It contains information on video surveillance technology, its application, shift to IP networks, leading-edge video analytic techniques applicable to it, its needs, the developments and trends in this field, the issues it raises, and the supply and demand it generates.




Video surveillance is a segment of the physical security industry, which also includes access control, fire detection and control, the technical management of buildings, systems to ensure individual safety and the detection of intrusion.




Video surveillance consists of remotely monitoring public or private places, using mostly power-operated cameras that transmit the images taken to monitoring equipment that records or reproduces the images on a screen. It captures images of moving people in order to monitor comings and goings, prevent theft, assault and fraud, as well as manage incidents and crowd movements.




 Video analytics, also called intelligent video surveillance, is a technology that uses software to automatically identify specific objects, behaviours or attitudes in video footage. It transforms the video into data to be transmitted or archived so that the video surveillance system can act accordingly. It may involve activating a mobile camera in order to obtain more specific data about the scene or simply to send a warning to surveillance personnel so that a decision may be made on the proper intervention required.


Intelligent video surveillance systems use mathematical algorithms to detect moving objects in an image and filter non-relevant movements. They create a database that records the attributes of all the objects detected and their movement properties. Decisions are made by the system or events of interest are searched in archived footage based on rules (e.g., if a person oversteps a boundary, send an alert).




      The video captured by surveillance cameras must be sent to the recording, processing and viewing systems. This transmission can be done by cable (coaxial or fibre optic cables, stranded copper wire) or by air (infrared signals, radio transmission).




      Wired video is the most predominant in video surveillance systems. It offers greater bandwidth and better reliability than wireless connections, at a lower cost. However, wireless video transmission is sometimes the best solution, for example when monitoring large perimeters where installing cables would be too costly, or when the areas to be monitored cannot be reached by cable.




      Whether wired or wireless, the video signal can be analogue or digital. Most video transmissions for surveillance are currently still analogue. However, computer networks (LAN, WAN or Internet) are increasingly used to send video using the IP protocol. IP cameras can be directly connected on these networks, whereas the video flow coming from analogue cameras must first be digitized by an encoder, also called a video server, in order to pass through the IP networks.




      Video management systems process video surveillance images, such as managing different video flows, and viewing, recording, analyzing and searching recorded footage. There are four major categories of video management systems.










Digital Video Recorder (DVR22):




 Device with an internal hard disk for digital recording of video and built-in video processing software. It accepts only flows from analogue cameras, which it digitizes. Recent models make it possible to view the video remotely on computer. Still quite widespread, it is slowly being replaced by systems that support IP video from end to end.




• Hybrid Digital Video Recorder (HDVR23):




Similar to the digital recorder, but accepts connections from both analogue and IP cameras. Several types of digital video recorders can be made hybrid by installing a software application.




• Network Video Recorder (NVR24):




 Designed for video surveillance IP network architectures, it can only process video signals from IP cameras or encoders.




• IP video surveillance software:




 Purely software-based solution for managing video on an IP network. For surveillance systems with few cameras, a Web browser may be enough to manage the video. For larger video surveillance networks, a dedicated video management software application must be used, which is installed on a PC or server. Although more complicated to install due to the required server configurations, it offers greater flexibility with respect to choice and the addition of video surveillance network parts. IP video surveillance software applications are a major trend in video management, especially in infrastructures with large numbers of cameras. Open platforms allow for easy integration of cameras and hardware components from different manufacturers.








The detection and tracking of faces and facial features in video sequences is a fundamental and challenging problem in computer vision. This research area has many applications in face identification systems, model-based coding, gaze detection, human-computer interaction, teleconferencing, etc. In this paper, we present a real time system that performs facial features detection and tracking in image sequences. We are interested in developing intelligent interfaces and this system is the first step to give computers the ability to look at people, which may be a good way to improve human-computer interaction.




We have used a color-based approach to detect the presence of a human face in an image. A Gaussian distribution model was generated by means of a supervised training, using a set of skin-color regions, obtained from a color face database. Basically, the input image is compared with the generated model in order to identify face candidates, i.e., skin-color regions. The presence or absence of a face in each region is verified by using an eye detector based on an efficient template matching scheme. There is biological evidence that eyes play the most important role in human face detection.






































1) To prevent theft from happening both indoors and outdoors of the premises. The storage facilities and delivery areas within multiplexes are easy targets of small time thieves who are lurking for some fast cash. With the right supervision you can monitor your staff and keep in control that only authorized personnel enter the office.




2) While watching a movie or playing a video game there are chances of it getting ugly over a feud. These security cameras are a deterrent to all the violent outburst and fights that might occur.




3) Every business is running in the world due to the strong customer base that it has, if they only are the prime target then the direct impact is to the business. Ensuring that your customers are your prime important entity and their security is well monitored is important.




4) Damage to your belongings can be done by a person you trust to catch him right in action a security camera is your ideal tool. It ensures work ethics among your staff.




5) we cannot undermine the importance of this footage that is captured on tape as this is quite valuable in court of law. This video of any culprit committing a crime is a proof enough to put him behind the bars.


























1.3       SYSTEM STUDY






1.3.1 Existing System






In the existing system, it is difficult to identify the movement of  persons from the recorded video. From the recorded video , we cannot find the movement of particular persons. It is easy to identify only the movement of all the persons. It is difficult of identify the tracking of unauthorized persons in an important place such as banks, etc.,




Hidden video cameras are concealed cameras that are used to observe a subject or area without anyone being aware of their presence. Hidden surveillance cameras come in all shapes and sizes, and with a wide array of available features.




Since they are meant to remain hidden from the observer, they are often made to look like objects that have other uses. For example, hidden cameras are often made to look like clocks, computer speakers, and smoke detectors. They are often small enough to be concealed in objects such as pens, cell phones, or even a button.




      The current system, uses only recording of videos. In the current system, tracking of a particular person is not done. It does not deal with checking whether a user is authorized or unauthorized.




















1.3.2 Proposed System




Tracking of humans is an important computer vision building block. This is needed for many applications, ranging from surveillance and military through computer-aided driving and advanced human-machine interfaces. It will be more useful for tracing a particular person in a video.




The detection and tracking of faces and facial features in video sequences is a fundamental and challenging problem in computer vision. This research area has many applications in face identification systems, model-based coding, gaze detection, human-computer interaction, teleconferencing, etc.




Though human beings detect/track faces with very little effort, it is not easy to train a computer to do so. In pattern recognition parlance, human face is a complex pattern. Different poses and gestures of the face accentuate complexity.




The detection scheme must operate flexibly and reliably regardless of the lighting conditions, background clutter in the image, multiple faces in the image, as well as variations in face scale, pose and expression. The system should be able to detect the face even under small occlusions.




For tracking, once face detection is performed for a certain frame the detected positions are projected into subsequent frames followed by skin analysis within the projected regions to locate face(s) in these frames.
















1.3.3 Feasibility System




Feasibility study of proposed system is carried out to observe how far it would be beneficial to the organization. The feasibility analysis depends on the initial investigation. The idea for changing originates in the environment or from within the firm on the problems is verified. Initial investigation is conducted to determine whether the changes are feasible. Depending on the result of initial investigation the survey is conducted to more detailed feasibility study.


A feasibility study is a test of system proposal according to its workability, impact on the organization, ability to meet the user needs, and effective use of resources. It evolves around investigation and evaluation of the problem, identification and description of system specification of performance and the cost of each system, final selection of the best systems.


Objective of the feasibility study is considered to be feasible, only if the proposed system is useful and is determined at the preliminary investigation stage.


Any project is considered to be feasible only if the proposed project is useful to the organization. In feasibility study we consider the economical aspect of the problem, which is being studied. Three key considerations that were involved in the feasibility analysis are Technical, Operational and Economical.


                    i.            Technical feasibility


During the technical analysis, it is found that .Net contains all the features that are required to create windows based application and it is a available licensed project in this  Organization. It is platform independent. So this will be technically feasible to build the system.


                  ii.            Economical feasibility


Economical feasibility mainly deals with the following steps.


  • Cost of the user system.
  • Maintenance cost of the system.
  • Other costs related with software and hardware.


As the project not needs, extra hardware and software, cost of the project is economically feasible.


                iii.            Operational feasibility


The operations of the system are so simple, so that a minimal knowledge of English is enough to understand the working of the system. Any one can work with the system very easily as it supports user-friendly approach. So it is operationally feasible.














CALL: 08985129129 ,  E-Mail Id: