SUSPICIOUS MOVEMENT DETECTION THROUGH COMPUTER VISION

In these days, technology is playing a vital role in the fields of education. Several technologies related ideas, systems, applications, and products have been developed to facilitate the daily routine tasks in the education domain such as, learning management systems, campus management systems, bio

Project Title

SUSPICIOUS MOVEMENT DETECTION THROUGH COMPUTER VISION

Project Area of Specialization

Artificial Intelligence

Project Summary

In these days, technology is playing a vital role in the fields of education. Several technologies related ideas, systems, applications, and products have been developed to facilitate the daily routine tasks in the education domain such as, learning management systems, campus management systems, biometric-based attendance system, camera-based attendance systems, and e-assessment systems. We believe that one of the most important and hectic job of teachers in educational institutes is to perform the invigilation duties vigilantly. Teachers need to be very active throughout the invigilation time and move around the class to affirm that no one is cheating in exams. Despite of so many efforts taken by the invigilators in exam halls, many students succeeded to using unethical means in exams. Moreover, several studies have reported that the education of Sindh has suffered a lot because of cheating in exams.  Therefore, there is a dire need to address this issue with the help of technology. This project aims to develop suspicious movement detection for students in exam hall through the help of computer vision and deep learning-based approaches. This project will use a Camera to detect students’  suspicious movements in exam halls. We will use the Frames to detect the movements of heads, and then in frames, we see the moment of the head, after that we will compare the frame. For that we will set the threshold, if the moment of the head is detected then our system notifies the suspicious detection to our invigilator. In addition, this project will also develop a mobile application that will help the invigilator track the suspicious students through red alert coming from system of students who are moving too much in exam halls. We believe that the proposed system will bring revolution in exam halls and reduce the chances of cheating in exam halls. Furthermore, it will assist effectively the invigilator in tracking the suspicious students in exam halls.so our education system will be free from cheating system.For the detection of the cheating implemented the Yolov3 algorithms, computer vision, and Deep learning will be used as backbone architecture. The details regarding architecture were discussed in the methodology. Checking ever single candidate manually creates discomfort, fatigue, and stress among employees. Besides that, this manual process affects many things such as cost of time, resources, budget, and efforts done by management.

Project Objectives

To ensure the principles and integrity of exams and to prevent cheating, a system based on computer vision is proposed in this project. It will detect cheating by the detection of HEAD  movements , eye movements  through a surveillance camera. It is more precise as compared to human labor. This system is better and more effective than the traditional invigilation system as it does not require as much labor, energy, effort, and time as needed in the conventional system.This manual process not only affects the productivity but also it affects the budget, time and resources in a way that to validate thousands of candidates they have to hire and pay extra employees, they have to put extra efforts and consume extra resources for validation of candidates. Moreover, when the density of candidates becomes too high, incidence of injury and illness, severe traffic delays, and pollution also increase, often more than proportionately through the interaction of populations. As densely populated areas are also ideal for the development and rapid spread of some respiratory epidemics.

Project Implementation Method

Human cheating behavior is very peculiar in nature. In the case of exam hall, it becomes tougher to detect the correct cheating behavior and bypass the false ones. Our major goal is to detect head movement from video input continuously whether the head is in movement or not. In this project, we will use a fake API to detect and track the head pose, We will check the angle of the head whether it’s in a normal position or not, we will set a threshold according to that we will compare frames, for example, 1st frame with 8th frame and it will check continue, if the system finds suspicious movement them it will send a notification to the invigilator, this system will also save a video of a particular part, invigilator can show that as a proof to the student.

We will use OpenCV, Tensor and yolov3

OpenCV

OpenCV (Open-Source Computer Vision Library) is an open-source computer vision and machine learning software library. OpenCV was built to provide a common infrastructure for computer vision applications and to accelerate the use of machine perception in the commercial products.

Tensor

Tensors are simply mathematical objects that can be used to describe physical properties, just like scalars and vectors. In fact, tensors are merely a generalization of scalars and vectors.

Yolov3

YOLO is an algorithm that uses neural networks to provide real-time object detection. This algorithm is popular because of its speed and accuracy. It has been used in various applications to detect traffic signals, people, parking meters, and animals.

Proposed model Detect cheating in a classroom during exams is the aim of the problem. For this, modified yolov3 has been used. It will detect cheating by the head and IRIS movements of students in the examination. The proposed model will considers cheating if students are looking around instead of detecting their HEAD movements. It considers students as no cheating who do not look around but do their paper as shown in Figure.

Benefits of the Project

• Effective and practical for high quality monitoring of online exams.
• Cheating will be reduced.
• There will be no injustice all students will be graded equally.
• After implementing this system and after deployment there will be more chances of merit quality and excellence.
• Every student will pay more concentration towards studies as he/she will have no chance for cheating.

Technical Details of Final Deliverable

The device of suspicious movement has some locations to be placed. Hardware for getting suspicious movement is pre-defined and cannot altered.

In this project, we will check the angle of the head, whether it is in average position or not. We will set a threshold according to that; compare frames, for example, first frame with eighth frame. It will check to continue. If the system finds suspicious movement, it will notify the invigilator. This system will also save a video of a particular part. The invigilator can show that as proof to the student. Software used in the android app. It will only keep the data and notify the invigilator when any suspicious movement occurs. A time interval is set to send the notification to the invigilator from time to time.

The proposed system will only provide information regarding universities, or any testing area of Pakistan, including international universities. The language of the interface, which interact in English.

The proposed system will only provide information regarding universities, or any testing area of Pakistan, including international universities. The language of the interface, which interact in English.

  The interface for android app is constant. Only android Operating system can used to run the app.

Student will recognizes the cheating done by the students in the classroom during exams. The system will send real time data and android app through camera. Invigilator will get an alert in the condition of suspicious movement of students.

FUNCTIONAL HIERARCHY

FR1: OPEN CAMERA

• Capture Frames

FR2: ALGORITHMS

• Identify Student Face, and assign ids
• Detect Head Movement
• If Head Movement will occur according to our threshold, then it will send notification to invigilator.

FR3: ANDRIOD APPLICATION

• Get Username of invigilator
• Get Password of invigilator
• If any suspicious movement detected, then he/she will receive notification.

FR4: STUDENTS

Software Used

Anaconda

Android (Mobile application)

Camera

Apache Web Server

Final Deliverable of the Project

HW/SW integrated system

Core Industry

IT

Other Industries

Education , Others

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Quality Education

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com