Project Name:
ESP32-Cam Object detection with Tensorflow.js
Team name:
Tanmay Soni
Vishal Kaushal
Siddharth Rathore
Vaibhav Songara

sdfjsdklfj




G

ESP-32 Object detection with Tensorflow.js

• Objective
• Introduction
• Literature Survey
• Proposed Methodology
• References

Objective:
The objective of this project is to develop a motion detection system using ESP32 CAM and TensorFlow.js to detect the presence of people in a room.

Introduction
The ESP32 CAM is a low-cost, low-power consumption module with a built-in camera. TensorFlow.js is a JavaScript library for training and deploying machine learning models in the browser. In this project, we aim to use ESP32 CAM and TensorFlow.js to develop a motion detection system that can detect the presence of people in a room.

Literature Survey
Previous research has shown that object detection can be achieved using various machine learning algorithms, such as YOLO (You Only Look Once) and SSD (Single Shot MultiBox Detector). TensorFlow.js has been used in previous projects for training and deploying machine learning models in the browser, but there is limited research on using ESP32 CAM and TensorFlow.js for motion detection.

Proposed Methodology
The proposed methodology for this project involves the following steps:

Capturing images of a room using ESP32 CAM and processing them to detect the presence of people using a machine learning model trained with TensorFlow.

Comparing the current frame with the previous frame to detect motion.

Using a threshold to determine whether the detected motion is significant enough to trigger an alert.

Deploying the motion detection system on a web server using TensorFlow.js.

References
Redmon, J., & Farhadi, A. (2018). YOLOv3: An Incremental Improvement. arXiv preprint arXiv:1804.02767.

Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C. Y., & Berg, A. C. (2016). SSD: Single shot multibox detector. In European conference on computer vision (pp. 21-37). Springer, Cham.

TensorFlow.js. (2021). TensorFlow.js. Retrieved from https://www.tensorflow.org/js

Application
Security system for homes or offices: The motion detection system can be used as a security system to monitor a room and detect any unauthorized entry or suspicious activity.

Automatic lighting system: The system can be integrated with an automatic lighting system to turn on lights when motion is detected in a room and turn them off when there is no motion.

Energy-saving system: The system can be used to reduce energy consumption by turning off lights and other devices when no motion is detected in a room.

Elderly care: The system can be used to monitor the activity of elderly people in a room and detect any falls or accidents.

Baby monitoring: The system can be used to monitor the activity of a baby in a room and alert the caregiver if the baby wakes up or moves around.

Retail stores: The system can be used in retail stores to monitor customer activity and detect suspicious behavior or shoplifting.

Healthcare: The system can be used in hospitals or care facilities to monitor patient activity and detect any falls or accidents.

Future Scope
Integration with other sensors such as temperature and humidity sensors to create a more comprehensive smart home system

Improving the accuracy of object detection by training the model with more data

Incorporating face recognition technology for more advanced security systems

Using multiple cameras to create a more comprehensive and accurate monitoring system