Multi Camera Adapter Module for Raspberry Pi

Published by Lee Jackson on

Before We Start

Standard RPi boards (not Compute Module) accept a single MIPI camera, but Arducam offers solutions that take multiple. This blog is the release of the first revision that supports 5MP V1.3 camera multiplexing in 2015.

Later, we’ve released a newer adapter to add 8MP support, and then a Stereo Camera HAT to disguise multiple cameras working simultaneously as a single connection acceptable to the Pi. Therefore, this product is officially end-of-life.


This Raspberry Pi multi-camera adapter module is designed for connecting more than one camera to a single CSI camera port on the Raspberry Pi board. Each adapter board can connect up to 4 cameras and the user can stack up maxim 4 adapter boards, which means up to 16 cameras on a single Raspberry Pi board.

Please note that this Raspberry Pi multi-camera adapter board is a nascent product which may have some stability issues and limitations because of the cable’s signal integrity and RPi’s closed source video core libraries, so use it at your own risk.


  • IoT cameras
  • Robot cameras
  • Wildlife cameras
  • 3D scanner


  • Accommodate 4 Raspberry Pi cameras on a multi-camera adapter board
  • Stackable and maxim cascade 4 adapter boards
  • 3 GPIOs required for multiplexing on one adapter board, 5 GPIOs for two adapter board, 9 GPIOs for four adapter board
  • All camera ports are FFC (flexible flat cable) connectors
  • DIP switches for easy stack up configuration
  • Support Raspberry Pi A/B/B+ and Pi 2.

Pin Configuration

DIP switches are used here for easy stack up configuration. When only one multi-camera adapter board is used, the switches 1and 5 should be switched to ON position. If two multi-camera adapter boards are used, the downside board should be switched 1 and 5 to ON and upside board should be switched 2 and 6 to ON position, and if 3 or 4 multi-camera adapter boards are used together each board layer should be configured as the Table 1 shown.

Table 1 Pin Configuration


For proper operation, only one camera should be enabled at a time. In case of only one multi-camera adapter board is used, driving Pin12 HIGH and driving Pin11, Pin7 to LOW to enable camera A on the adapter board. Enable camera B, C, and D, please refer the Table 2 for camera selection configuration. To disable all cameras on one adapter board, Enable 1 and Enable 2 signals should be toggled HIGH. Care should be taken that it is not allowed to drive the Enable 1 and Enable 2 LOW at the same time. When connecting more than two multi-camera adapter boards, only one of the cameras can be enabled by enabling the desired camera on one adapter board and disable all the cameras on the rest of adapter boards.

Table 2 Camera Selection Configuration

Camera Selection ConfigurationSelectionEnable 1Enable 2
No CameraX11

(X: don’t care)

Software Demonstration

Before using it, required dependency python-rpi.gpio have to be installed with

sudo apt-get install python-rpi.gpio

the command from the Terminal. Then run the following code for the quick demo.

import RPi.GPIO as gp
import os


gp.setup(7, gp.OUT)
gp.setup(11, gp.OUT)
gp.setup(12, gp.OUT)

gp.setup(15, gp.OUT)
gp.setup(16, gp.OUT)
gp.setup(21, gp.OUT)
gp.setup(22, gp.OUT)

gp.output(11, True)
gp.output(12, True)
gp.output(15, True)
gp.output(16, True)
gp.output(21, True)
gp.output(22, True)

def main():
    gp.output(7, False)
    gp.output(11, False)
    gp.output(12, True)

    gp.output(7, True)
    gp.output(11, False)
    gp.output(12, True)

    gp.output(7, False)
    gp.output(11, True)
    gp.output(12, False)

    gp.output(7, True)
    gp.output(11, True)
    gp.output(12, False)

def capture(cam):
    cmd = "raspistill -o capture_%d.jpg" % cam

if __name__ == "__main__":

    gp.output(7, False)
    gp.output(11, False)
    gp.output(12, True)

As normal camera operations, users can still use raspistill or raspivid command to take photos or videos. By control the GPIOs according to Table 1 and Table 2 configuration, users can take photos or videos by switching between different cameras. Have fun!

Categories: Applications


Jeremy · March 17, 2016 at 5:16 pm

Do you sell the adapter board?

Bruce · January 27, 2017 at 5:57 pm

How can I buy this module ? I’m interested to have one.

Mathieu · June 14, 2019 at 9:26 am

I would like to know if this multi-camera module is also compatible with Jetson Nano products ?
( or if it has to be connected to a raspberry pi ?


Ravi · December 8, 2019 at 11:48 am

Hi, I would like to buy camera adapter for raspberry pi 4. Can you please confirm the availability. Thank you

Keith · July 10, 2020 at 7:28 pm

I notice that it says “one camera at a time”. I want to connect as many cans as possible (i.e. 16) and take a simultaneous picture with all of them. My application is slow moving, do I might be able to take sequential pics from all the cameras if switching time is small: ideally not much more than about 5 milliseconds, based on getting all 16 images in less than 100 milliseconds. Does anyone know if that’s likely to be possible?

Philipp · July 22, 2020 at 12:10 am

Hello, I need a cable length of 70 feet between the cameras and the Raspberry. Will that work with the multi-camera adapter? I can use Cat 7 cable for extend.

    [email protected] · July 22, 2020 at 5:46 pm

    The signal integrity degrades a lot at 70 feet, and we’ve only had the extension tested with 10 meters (about 33 feet) HDMI cable

Diomedes Dominguez · August 2, 2020 at 7:11 am

@Keith, You will need to use the USB ports or use the approach of using IP Cameras.

Julius Pinsker · August 18, 2020 at 12:57 am

How fast is the switching latency? Would it be possible to quickly alternate between camera inputs to feed the made pictures into an obstacle avoidance ai?

Donie · September 8, 2020 at 8:33 am

Is it possible to buy this module?

jasfree · September 11, 2020 at 5:45 am

The Cameras and module requires power. Will the normal raspberry pi power adapter work (5v 3000mA)?

multicamuser · September 16, 2020 at 9:11 am

This board is so much more user-friendly. The i2c requirements on the newest boards tend to throw errors if the environment isn’t perfectly set-up. I just spent a couple of hours trying to get the v2.1 board to work with the RPi4 to no avail, then within three minutes got this board up and running successfully.

I wish this design had been maintained.

Leave a Reply

Your email address will not be published. Required fields are marked *