Vision based human tracking and localization for security surveillance

Project Title

Vision based human tracking and localization for security surveillance

Project Area of Specialization Artificial IntelligenceProject Summary

The main purpose of our project is to increase security by recognizing people and tracking them if they are unknown. Because in our daily lives, only standard security cameras are used, which have no record of any person, known or unknown, who enters a location. Previously, the record was kept by handwriting. In contrast, in our project, we can easily keep a record (entry and exit time with image) of an unknown person as well as track an unknown person.

Project Objectives

The main purpose of our project is to fulfill the security need which acquire less human effort and provide high level security. Our system will recognize all people and will track the unknown persons without any external help and also maintain the record of every person which enters in our society. Our system will have all the data like at which time the unknown person was present in our society. If any unknown person enters in our society or university for many times then an alert will appear to authority and if we want to know whether or not a person has ever entered our society, we can readily verify using our system. If any criminal is hiding in our society we can locate his position by seeing his track.

1. Identify the persons entering the site premises

2. Maintain the record of entrance of both known and unknown persons

3. Track and locate the position of the unknown persons in the project site.

Project Implementation Method

First of all, we will use cameras for making real time video of different location points. Then we process the videos using Machine learning and Deep Learning algorithms. First we will apply the machine learning algorithm to detect the human in the video. And then we’ll use the algorithm to detect the face of person. Then we apply the deep learning to recognize the person’s face. After recognizing the face, we will check from our dataset that it is present in our record or not. If record founded it means it is known person otherwise it is unknown person. If it is unknown person, then we label that person and after that tracking will start. In tracking when the person comes in front of next camera, again detection and recognition process takes place and checked that person is labeled or not. If it’s labeled, the record will update (camera no, and time will add in his record).

Benefits of the Project

The benefit of this project is as follow:

1. It increases the safety of the site. 
2. it maintains the record of the person who enters our building.
3. The alert sends to authorities when an unknown person enters the building.
4. The record shows how many persons are present in the building.
5. The record also shows how many unknown persons are present in the building.
6. Track record of an unknown person is maintained in this project which shows where an unknown person went in our building. 
7. Unknown person can go only to a specific area in our building and the remaining building only allows for a known person.
8. All door of our building is operating automatically.

Technical Details of Final Deliverable

The technologies which we use in this project are given below:

1.  First of all, we use concepts of machine learning and build the sketch of our project.
2. Using the TensorFlow library to train our model of face recognition.
3. Use the face recognition library for recognizing the persons who enter the building.
4. Use IoT technology and integrated it with machine learning to open and close the door.  

First of all, we train the model to recognize the known persons using the TensorFlow library on GPU. We take real-time video using cameras that are directly attached to the computer system. GPU process these images of the video and classify the known and unknown persons. After recognizing the person it decides either to send the mail or not to authorities. When this procedure has been done the system sends the signal to the IoT processor to open the door. Which door will open is decided on the bases of the signal sent to the IoT processor.  

Final Deliverable of the Project HW/SW integrated systemCore Industry SecurityOther IndustriesCore Technology Artificial Intelligence(AI)Other Technologies Internet of Things (IoT)Sustainable Development Goals Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	53150
Graphic Card	Equipment	1	40000	40000
camera	Equipment	5	1600	8000
ESP 32	Equipment	1	1000	1000
data cable	Equipment	1	200	200
card board	Miscellaneous	1	1000	1000
Saw	Miscellaneous	1	200	200
convertor	Equipment	1	200	200
4 in 1 USB connector	Equipment	1	400	400
servo motor	Equipment	5	270	1350
flex print	Miscellaneous	1	500	500
liquid Glue	Miscellaneous	10	20	200
Board Marker	Miscellaneous	2	50	100