In today’s world, there is an irregular increase in the crime rate and also there is an increase in the number of criminals, which leads to a great fear about security issues. Criminal detection using face recognition is the most leading technology to combat crime issues.
Crime anticipations and criminal identification are the main issues before the police personnel because property and life protection are the elementary fears of the police but to combat the crime, the accessibility of police personnel is limited. With the arrival of security technology, cameras particularly CCTV have been installed in many public and private areas to give scrutiny activities.
The footage of the CCTV can be used to recognize suspects on the scene. In this article, an automated system of facial recognition for the criminal catalog will discuss. This system will have the ability to detect faces and identify faces automatically in real-time. This will help the law implementations to detect or identify suspects of the case if no thumbprint present on the scene.
Criminal detection using face recognition
The face plays a central role in human identity. It is the feature that best differentiates a person. Face recognition is an exciting and stimulating problem and influences important applications in many areas such as an ID for law enforcement, verification for banking and security system access, and identification of personal among others. Over the years, most security methods have been developed that help in keeping private data secured and restraining the chances of a security breach. Face recognition is one of the few biometric methods that own the merits of both high precision and low indiscreetness is a computer program that uses a person’s face to robotically identify and verify the person from any digital image or a video frame from any video source. It compares designated facial features from the image and a face database or it can also be hardware used to validate a person.
Aim of this technology:
The aim of Criminal detection using face recognition is:
- Matching a face with the existing database precisely.
- Applying principal module analysis for finding unique features from many images to get the resemblance for the target image. The remaining of this paper is planned as follows
System of face recognition:
Face Recognition for identification Criminals is a system in which the investigation team will input an image of the person in query inside the system and then the system will first preprocess the image which will cause unsolicited elements for example noise to be eliminated from the image. After that, the system will then categorize the image on the basis of its benchmarks such as the length between the eyes, the distance of the jawline, etc. Then, the system will run an exploration through the database to find its flawless match and show the output. This work is emphasizing on applying the system for criminal identification
Face detection is the primary step in making a facial recognition system. This is where the system detects the face and knows whether it is certainly a human face or not. It also makes sure whether the system can differentiate between the subject and the background hence letting it detect and identify faces easily.
The design and development of a face recognition system
The design and development of a face recognition system consist of five stages which are
- Requirement analysis
- Implementation and testing
- And maintenance
The planning stage is where the system is being deliberated i.e., why and how the system will be developed It is alienated into two phases which are as follows:
- Initiation of Project – an initial analysis is assumed about how to assemble face images to be used as the model to the system.
- Planning of project– Here we will determine the correct technique/ software to be used for the detection and recognition.
Requirement analysis defines the analysis that is needed for developing the planned system by functional and non-functional necessities. Functional requirements plan what the system should do and support the worker’s activities.
Here is the list which shows the functional needs for the face recognition system
- Firstly, The system lets the user login by using the username and password given which is default as “admin”
- Secondly, It lets the user to the input image be compared.
- Thirdly, The system lets the image be matched
Design System project defines the manner, components, modules, interfaces, and data for a system needs.
According to this illustration, the first step is to develop face databases which is the matching template for the system. A face database is formed by getting a collection of people’s photos. The photo must be half body photo where the face is facing visible. In the process of id verification for an image, the image which is taken using a digital camera would be processed. The image will be distinguished and extracted and ready for the subsequent stage. The next stage is pre-processing of the image, where redundant features are removed. This is to reduce needless processing effort.
In the extraction feature, the images are accumulated from the database and signify it as a vector, then the algorithm will find the normal face vector or the mean and it will subtract the mean face from each template face. All these photos then are managed as the standard features of the human face. These features will be used in the recognition phase where it tries to match with the correct image in the database. If the image will be matched, then the identification of the image will be confirmed, otherwise, it will stop
Implementation and testing:
The implementation and testing phase of the criminal detection using the face recognition system includes implementation on the backend and on the interface coding. The system interface was executed using Microsoft Visual Studio while the backend components, which are database and coding, were applied fully using MATLAB R2013b.
Function OutputName= recognition (test image, m, A, Eigen faces)
Function OutputName is used for the recognition of images. This function has 4 parameters which are the TestImage, m, A, and Eigenfaces. TestImage is the input image that we need to find the corresponding image in the databases. m is the mean image in the database and A is the nonconformity of the images. Lastly, Eigenfaces is actually the Eigenvectors of the covariance matrix of the working database
Input Image= Imread (test image);
Temp= Input image, (:, :, 1);
(irov, icol)= size (temp);
In image= reshape (temp, irow*icol, 1);
Difference= double (Inimage)-m; % centering the image
This shows phases in extracting all features for matching actions. Line 1 of the function will speak the input image. Line 2 shows that a 1-dimensional momentary file of the input image is formed. 3rd Line is used for setting the size of the file of the image. It is to be remembered that the dimensionality must be the same for all images. Line 4 will redesign the image by m x n resolution and then the face characteristic is extracted from the input image. After, it will ready for ID procedures.
How this technique is helpful?
This unit highpoints the main consequence of the face recognition system and its advantages. This study intends to project, develop and test the Face Recognition system for Criminal Identification. The main function of a face recognition system is image identification which was programmed with detection and extraction of images, projecting images,s, and recognition of the image. The user needs to input the image for the recognition process. Once the image is identified, detected, and extracted, all the essential features are removed for identification
Overall, there are numerous advantages that have been recognized as follows:
- Firstly, A face recognition system is a better alternative for criminal identification as a substitute for using thumbprint identification.
- Secondly, It can easily automate many identification activities. For example, a criminal photo was taken through CCTV just needs to enter into the system for identification. The system will then run robotically from identifying, detecting, and extracting the image, features extraction, and identification actions.