Esp32 camera webserver for data collection

ฉบับไทย

This project is dedicated for collecting images from Esp32S3 to train FOMO object detection model in Edge Impulse. The Hardware that was used were provided by Wireless Solution Asia. Users can collect images from the webserver and download to their computer directly. User can aslo manipulate Esp32S3 camera resolutions and sensors settings. Just unzip the project inside the Arduino directory and you're good to go!

What you'll need

• Arduino IDE, preferably the latest ones, but older versions will still do the job.
• ESP32-S3 is preferable but older version will do just fine.
• OV2640 camera or any OV series.

Project files descriptions

1. camera-webserver-for-esp32S3.ino - Containes Arduino codes for streaming camera images to the webserver.
2. index.html - Contains HTML elements for streaming.
3. index.js - Javascript functions for HTML element and streaming.
4. styles.css - makes webserver looks more lively.

How to Install and run the project

1. Download the project and unzip it to your Arduino Directory.





2. Open camera-webserver-for-esp32S3 directory and launch camera-webserver-for-esp32S3.ino. Enter WIFI SSID and PASSWORD.





3. Under tools change your Board to "ESP32S3 Dev Module" and PSRAM to "OPI PSRAM".





4. Upload the code to your ESP32S3 and copy the IP address.





5. Open index.html file, and paste the IP Address obtained from previously.





6. Done!!!

- I've created repository to provide simple guide to training FOMO object detection model please visit train-FOMO-object-detect-esp32.

Webserver features

• User can switch languages on the top right corner of the web.
  
• Adjustable Camera Settings, to see more details about each setting please visit https://heyrick.eu/blog/index.php?diary=20210418&keitai=0.
  
• User can hide the setting console by clicking on the icon above.
  
  
  
  
  
• Adjustable Resolutions.
  
• Class labels will give each image an assigned name when downloaded, suitable for labeling images or organizing data collections directories.
  
  
  
  
  
• Capture button will start capturing images from stream and display it in the gallery below.
  
• The webserver can capture each images in an interval according to the input value. The interval cannot be less than 0.1.
  
• Set Instances to limit the amount of images captured.
  
  
  
  
  
• Download button located on the top left will zip all captured images into a single file and download.
  
• Clear buttons located on the top right will delete all captured images.
  
• User can aslo delete each image individually.
  
  
  
  
  
  

Credit

Thanks to WIRELESS SOLUTION ASIA CO.,LTD for providing AIOT board to support this project. Also thanks to [RandomNerdTutorials](RandomNerdTutorial for providing essential codes for streaming Esp32 camera to webserver.