A profiling frame work using knowledge graph
In this project, we are going to present research work about profiling humans under on-campus environment using face recognition we have to collect data through face recognition by using raspberry pi and pi cam that will send the current timestamp, date, and camera location after some time then
2025-06-28 16:30:06 - Adil Khan
A profiling frame work using knowledge graph
Project Area of Specialization Artificial IntelligenceProject SummaryIn this project, we are going to present research work about profiling humans under on-campus environment using face recognition we have to collect data through face recognition by using raspberry pi and pi cam that will send the current timestamp, date, and camera location after some time then we will use this data to identify the location in real-time plus autonomously. Further on the top of this data we can make a knowledge graph and make a ruled base querying engine. Raspberry pi connects with the camera to run facial algorithms on it for the capture video feed. For example, an application related to this is Employees Monitoring application Industry track their employees through it at what time where the employee was present, we present a framework for profiling university students/members in our university setting using the camera feeds with the in cooperation of IoT and semantic web, semantic and profiling information is key to the successful implementation of it.
To understands the importance of data We need semantics in this work we have proposed a semantic framework which integrates semantic information of student/worker through knowledge graph, temporal profiling information and facial recognition using the proposed framework user can answer semantic quires of the student/worker, working in the smart university It aims at supporting intelligent quires by using knowledge graph. Its uniqueness is that it doesn’t leave data preprocessing it and then makes a knowledge graph and built a ruled base query engine on it. It is helpful to suspect where the human is and where he was to be at that time, for example, an accident of short circuit takes place in any industry and the man there for safety duty was not present there he was somewhere else due to his negligence towards his duty make the company loss so by this tracking we can easily cope with unhandled situations and keep track of all things.
Profiling, the act of suspecting or targeting a person on the premise of discovered characteristics or behavior or the recording and analysis of a person's psychological and behavioral characteristics.
The Knowledge Graph is used by any farm it contains information and data and it can help in delivering more accurate results to search engines in other words a knowledge graph is a programmatic manner to version an information area with the assist of subject-rely experts, statistics interlinking, and gadget gaining knowledge of algorithms.
Our project objective is that can we make profiling possible with lightweight IoT based systems (pi camera with raspberry pi).
Project Implementation MethodIn our project, first we are going to collect data so for data collected we used face recognition. We want data for making the profile framework which consists of a knowledge graph, query engine, and semantic web. So, the base of our project is data, what kind of data? The data of multiple persons how can we collect that data in a campus environment? We collect this data by face recognition this was our first and basic step of this project without data we can’t do profiling. So, it is a must for us to collect a sufficient amount of data to build up this project. For data collection, we go towards the IoT side because nowadays it is one of the most emerging fields and this helps in developing intelligent systems using intelligent tools. We used Raspberry Pi and Pi camera module v2 as smart tools IoT Remote access and different device controls are allowed as part of information systems. We try our best to use the latest technologies as, so the language that supports this project is Python, let us discuss the main method of data collection this is the main motive we try two algorithms implementations of face recognition named frontal face using OPEN CV HOG method through it we can collect our data after that we store this data into CSV file because data was consisting of multiple values including Name, Id, and Timestamp and Camera location. Name, the id of the person detected on the camera and the time at which he/she was detected by the camera, and the location at which the camera has placed the values of these parameters of data we have after going through facial recognition methodology this should we have to achieve.The system's output is based on three phases, which are python script datasets, trainer, and detector. OpenCV is an algorithm used for image recognition, especially for face detection. , the face recognition technology would not exist without facial data collection. It is indeed an important part of the whole process. With the help of this technology these automated systems can be used to check or identify the identity of personalities in just a few seconds based on their facial features. After data collection we are going work on knowledge graph construction and the top of it query engine construction have been done.
The knowledge graph is proposed as generic for usability in any industry and enterprise for profiling purposes. Basic working on (Knowledge graph, query).
Data preprocessing it and then makes a knowledge graph and built a ruled base query engine on it. It is helpful to suspect where the human is and where he was to be on that time.
There are many benefits of our project including real-time tracking of entities from the real-world environment, fast and Non-Invasive Identity Verification, monitoring of students/employees and it makes profiling feasible with the technology of face recognition and embedded processing the capabilities of portal devices such as the Raspberry Pi can greatly help law enforcement agencies to identify suspicious activities in a cost-effective, intelligent manner safety. The Raspberry Pi does not only provide the right fit feature detection and extraction platform. The proposed framework automatically scans the face using a frontal face algorithm, followed by HOG as feature extractor and Haar as a classifier.Also the cooperation of IoT and semantic web, semantic and profiling information easy integration of semantics with profiling data and integration of querying on the top of knowledge graph.
Technical Details of Final DeliverableWe have proposed a human profiling model that uses facial recognition to obtain data in an on-campus environment and then perform further analysis on the data collected from the face recognition technique in face of profiling framework using this data and knowledge graph.
facial recognition used for data collection and on collected data knowledge graph was built up and further on that knowledge graph and semantic web for querying.
the importance of data We need semantics in this work we have proposed a semantic framework that integrates semantic information of student/employee through knowledge graph, temporal profiling information, and facial recognition using the proposed framework user can answer semantic quires of the worker, working in the smart industry It aims at supporting intelligent quires by using knowledge graph. Its uniqueness is that it doesn’t leave data preprocessing it and then makes a knowledge graph and built a ruled base query engine on it and final deliverable it as a dashboard for profiling framework.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 18100 | |||
| Raspberry Pi Model 4B | Equipment | 1 | 11500 | 11500 |
| Pi camera v2 | Equipment | 1 | 4200 | 4200 |
| Power Supply USB-C | Equipment | 1 | 550 | 550 |
| HDMI to VGA | Equipment | 1 | 350 | 350 |
| Micro SD card | Equipment | 1 | 550 | 550 |
| Ethernet cable | Equipment | 1 | 200 | 200 |
| Heat sinks | Equipment | 3 | 250 | 750 |