Esp32 Wearable Ecg

A Benchmark Reference For Esp32-Cam Module

1st Nowroz, Sayed T.

Dayton, Usa

2nd Saleh, Nermeen M.

Dayton, Usa

Abstract—The ESP32-CAM is one of the most widely adopted open-source modules for prototyping embedded vision appli- cations. Since its release in 2019, it has gained popularity among both hobbyists and professional developers due to its affordability, versatility, and integrated wireless capabilities.

Figure: System Model & Simulation Flow for Esp32 Wearable Ecg

Despite its widespread use, comprehensive documentation of the performance metrics remains limited. This study addresses this gap by collecting and analyzing over six hours of real-time video streaming logs across all supported resolutions of the OV2640 image sensor, tested under five distinct voltage conditions via an HTTP-based WiFi connection. A long standing bug in the official Arduino ESP32 driver, responsible for inaccurate frame rate logging, was fixed. The resulting analysis includes key performance metrics such as instantaneous and average frame rate, total streamed data, transmission count, and internal chip temperature. The influence of varying power levels was evaluated to assess the reliability of the module.

Figure: System Model & Simulation Flow for Esp32 Wearable Ecg

Index Terms—Internet of Things (IoT), Embedded Systems,

I. Introduction

Unlike general-purpose computers, embedded systems are optimized for real-time performance with power, speed, and memory constraints. Over the past decade, the rise of Internet- connected devices has transformed the consumer market, driven by advances in low-power computing and wireless networking . Portable IoT applications exist across diverse sectors such as industrial automation, robotics, smart agricul- ture, space and deep-sea exploration, and healthcare. These systems are significantly enhanced with the integration of cameras, enabling advanced imaging and analysis in scenarios where traditional infrastructure-based computer vision systems may not be feasible. Recently, the demand for cost-effective and vision-enabled embedded systems has increased.

Figure: System Model & Simulation Flow for Esp32 Wearable Ecg

In response to growing demand for affordable embed- ded vision solutions, devices like the ESP32-CAM have be- come widely adopted. This open-source module combines an OV2640 image sensor with an ESP32S microcontroller, offering compactness, wireless connectivity, and sufficient processing power to support IoT-based vision applications . Despite its popularity, there is lack of documentation on performance benchmarks and thermal characteristics of this hardware. This paper addresses that gap by presenting a comprehensive experimental study with real data, evaluating key performance metrics of the ESP32-CAM. The study incorporates findings enabled by a recent bug fix related to incorrect frames per second (FPS) calculation in the Arduino ESP32 Core repository, ensuring the most accurate and up-to- date assessment of the module’s capabilities. The contribution

Of This Study Is As Follows:

1) Bug Fix: A long-standing bug in the official ESP32 Arduino Core repository was resolved, which had caused inaccurate FPS reporting since 2018 .

2) Public Release of Benchmark Dataset: The authors published dataset containing timestamped video stream- ing logs (including FPS, temperature, and payload met- rics) collected at 15-minute intervals in all test cases, available at GitHub .

3) Comprehensive Benchmarking: An exhaustive eval- uation of frame rates in all supported resolutions of the ESP32-CAM was performed, establishing a detailed performance baseline for future design.

Ii. Background

The ESP32-CAM is a compact and low-cost development board introduced by AI-Thinker in 2019. It integrates the ESP32S microcontroller, built-in WiFi connectivity, a microSD card slot for local storage, an LED flash, and multiple GPIO and peripheral interfaces operating at 240 MHz. The board typically interfaces with the OV2640 image sensor, a 2- megapixel camera capable of streaming video and capturing still images in various resolutions. The ESP32S microcon- troller on board configures the image sensor via the I2C protocol, while the DVP interface delivers the encoded image in an electrical signal. Supported encodings include RGB565, RGB555, and JPEG, with JPEG compression enabling higher- resolution image capture (up to 1600x1200) at lower frame rates. In contrast, RGB formats are limited to lower resolutions (320x240) due to RAM constraints .

The Arduino framework is an open source platform de- signed to simplify embedded system programming through a rich ecosystem of hardware libraries and cross-platform compatibility. It uses a high-level language based on C/C++ and provides developers with tools to compile, and upload firmware to various microcontrollers . For ESP32-CAM, it exposes high-level APIs from the Espressif IoT Development Framework (ESP-IDF), streamlining integration of WiFi, Blue- tooth, and camera functionalities . Its accessibility makes it well-suited for rapid prototyping, educational projects, and lightweight commercial deployments.

Node-RED is an open-source, flow-based development tool built on Node.js, widely adopted for designing event-driven IoT applications . Its visual programming interface allows for connecting devices, APIs, and services using drag-and- drop nodes, minimizing the need for traditional coding. In this work, Node-RED is used to manage UART-based real- time video stream logs, perform string processing, and export data to CSV files. Its lightweight design and flexibility make it well-suited for real-time data handling in embedded system workflows like those involving the ESP32-CAM.

Iii. Related Work

The ESP32-CAM is a low-cost microcontroller with an integrated camera that has widespread adoption in domains such as agriculture, surveillance, healthcare, home automation, and more. Its open-source ecosystem and built-in WiFi have enabled rapid prototyping and deployment of vision-based systems for real-time monitoring. This section summarizes recent applications of ESP32-CAM across several sectors.

ESP32-CAM has been extensively used in agriculture au- tomation due to its compact form and imaging capabilities. In , it was used to continuously monitor poultry health by integrating machine learning and Internet of Things (IoT), providing early diagnosis using image-based data. Similarly, presented a solar powered ESP32-CAM system for remote plant growth monitoring. It has also been applied in automated weed detection and elimination systems using image-based ML interface . ESP32 has also been deployed for farm security to detect animals and human activity via pose estima- tion algorithms to detect intrusions .

In IoT architectures, ESP32-CAM often serves as an edge device, transmitting image data to local or cloud-based AI systems for analysis. In , a wearable ESP32-CAM system for Alzheimer patients is proposed, streaming video to a Raspberry Pi running MobileNet and DenseNet models for face detection and recognition. A similar approach is seen in , where the module was used to detect citrus diseases using MobileNet. IoT-based home security systems have made use of ESP32-CAM, using facial recognition for real-time surveillance , . The live video feed was streamed, triggered by human presence detection, with an immediate notification to the owner. Other IoT-based home security work uses real-time facial recognition for automated door access , recognizing visitors, and notifying the homeowner with photos of unfamiliar faces.

Video Frame

Fig. 1. Experimental setup and implementation for real-time video streaming log collection. ESP32 has enough resources to host lightweight AI models, optimized to perform in real-time with low power and memory footprint. has deployed variations of fully connected neural networks in the Xtensa core. In , a model is implemented to accurately detect flood, fire, traffic accident, and normal conditions on 96x96 images with 850 ms inference time. Applications such as robotic face tracking and aerial disaster detection in drones further showcase the potential of ESP32-CAM in TinyML use cases. ESP32-CAM has also been adopted in commercial products. Prusa, a very well- known company in the 3D Printer market, uses ESP32-CAM for real-time print monitoring . The trend in ESP32-CAM and vision-based IoT applications has been increasing almost exponentially since 2020. Multi-node IoT vision network and security have been another new trend in the field .

Iv. Methodology

In this study, 15-minute video streaming logs were collected for each supported resolution of the ESP32-CAM module, with video streamed over WiFi via an HTTP server. A client can access the HTTP server hosted by ESP32-CAM on the local WiFi network, by accessing the IP address of the ESP32-CAM module. Real-time logs were transmitted from the ESP32- CAM through the UART serial interface and collected via USB. Node-RED was used to access, parse, and write logs into CSV files. For each transmitted video frame, a comma- separated log string was generated containing real-time statis- tics, including instantaneous FPS, average FPS, runtime, bytes transmitted, and chip temperature.

The firmware used is based upon Arduino’s CameraServer example code, with corrected longstanding FPS logging bug in the example code. The firmware was modified to stream extra parameters such as temperature within the logs and dynamically control image resolution over UART. However, the resolution can also be changed from the HTTP interface.

To study the impact of power variation, the same approach was adopted with controlled voltage levels applied directly to

Related Journal Articles & DOI Links

Selected peer-reviewed publications relevant to 12 Lead ECG Acquisition. Click the DOI to access the full paper (may require institutional access).

Why Choose Us?

Bangalore guidance for robotics, Spectre and autonomous systems projects.