MicroPython | 2 min | 8908
This tutorial extends the article MicroPython: Taking photos with an ESP32 and in this case, the ESP32-CAM running MicroPython takes a photo every 5 seconds and saves it on the microSD card. The photos can be then combined into a time-lapse video! :)
The MicroPython official version doesn't support I2S interface. However, tsaarni wrote a driver for the OV2640 and forked the MicroPython repository to include it. Thus, you need to follow the instructions in this article: MicroPython: Taking photos with an ESP32 to install the right firmware.
Hardware and Software
In the following table, you will find the hardware and software that you can use on this tutorial:There are two new articles about this topic:
After flashing the MicroPython firmware, clone the repository
git clone -b timelapse-camera https://github.com/lemariva/uPyCam.git
then rename the file
boot.pyand configure the Wi-Fi access settings:
# wlan access ssid_ = "" wpa2_pass = ""
The ESP32-CAM uses Wi-Fi to connect to an NTP-server and to get the actual time. Additionally, you can check, if everything is working because the code enables an FTP server.
Then, connect your ESP32-CAM to a 5V battery and enjoy the results! :)
If you need a datasheet of the ESP32-CAM, here is one available, I needed to define the pins that are connected to the MicroSD slot.
Error on saving photos
Some of the photos were saved with error on the MicroSD card, and I got an error when I tried to export the video sequence. So, I wrote some code lines to repair the photos. The error that I got was the following:
Premature end of JPEG file.
Create a file
photo_repair.pywith the following code:
import os import cv2 import shutil from os import listdir from os.path import isfile, join basedir = os.path.abspath(os.getcwd()) onlyfiles = [f for f in listdir(basedir) if isfile(join(basedir, f))] new_folder = "numerated" for idx, filename in enumerate(onlyfiles): base, extension = os.path.splitext(filename) if extension == ".jpg": print(filename) img = cv2.imread(filename) cv2.imwrite(filename, img)
Then, install the dependencies (opencv) using pip:
pip install opencv-python
and run the code typing
This article presents an example code to use the ESP32-CAM as a time-lapse camera using MicroPython. The module takes a photo every 5 seconds and saves it on the SD. Then, using x264vfw and VirtualDub is possible to generate a time-lapse video. Some of the photos are saved on the microSD with errors and I included a some code lines to repair them.