E-Lock Demo Guide with the MCU-based Solution for Face Recognition

Showing results for 
Search instead for 
Did you mean: 

E-Lock Demo Guide with the MCU-based Solution for Face Recognition

No ratings

E-Lock Demo Guide with the MCU-based Solution for Face Recognition

This guide is intended as a reference for creating a demo application using the SLN-VIZN-IOT kit. In this guide, we will be constructing a demo e-lock application using the SLN-VIZN-IOT kit for secure face recognition using liveness detection/anti-spoofing. If you haven’t already, be sure to check out the Getting Started Guide for the SLN-VIZN-IOT kit here.

Build Process

Our e-lock design will make use of GPIO_AD_B0_2 and GPIO_AD_B0_03 to drive an H-Bridge circuit which actuates a lock using a 9-volt battery. These pins (and our ground) can be found on the serial header located on the front of the kit as shown below:


To build our e-lock, we will be modifying the sln_vizn_iot_userid_oobe application found in the SLN-VIZN-IOT SDK. Instructions for downloading the SDK and importing the userid_oobe application can be found in the ‘Get Software’ and ‘Build and Run’ sections of the Getting Started Guide.

The following video shows the modifications necessary to implement the E-Lock demo using the sln_vizn_iot_userid_oobe project

To enable these pins as GPIOs, we must modify pin_mux.h and pin_mux.c found under the board folder. For simplicity, we contained these initializations in a function called BOARD_InitDoorLockPins. The code to enable these pins was generated using MCUXpresso’s integrated Config Tools, although this is not necessary. The MCUXpresso Config Tools can be read about in-depth here.


Next, we need to make sure that the BOARD_InitDoorLockPins function we just created actually gets called so that the GPIOs will work the way we want them to. To do this, we will add the function call inside of our main function in main.c.


After adding the door lock initialization to main, we will modify sln_system_state.cpp found under the source folder to add the code which will toggle the GPIO’s we setup in the previous step. To do this, we will make use of the GPIO_PinWrite function found in “fsl_gpio.h.” Using this function requires us to add the line “#include fsl_gpio.h” at the top of sln_system.cpp like shown below:



The GPIO_PinWrite functions here will be used to unlock the door whenever a face is recognized (sysStateDetectedKnownUser) and lock the door whenever no known users are in view of the camera (sysStateDetectedNoUser).

With the software modifications complete, we need to compile the code and flash our kit with the updated firmware. This can be done by using the ‘Debug’ option found in the Quickstart Panel as shown below. Make sure that the project is compiled and flashed is the sln_vizn_iot_userid_oobe project by verifying the name of the project shown at the top of the Quickstart Panel.


For more detailed instructions about flashing the SLN-VIZN-IOT, check out the Flash and Debug SLN-VIZN-IOT Project section under Build, Run in the Getting Started Guide

With the software modifications complete and the updated firmware installed, all that’s left to do is to add some wires from the GPIO pins to the door lock and power on the kit.


Now our e-lock is ready to go!

When a user with an unrecognized face (indicated by a red LED) tries to turn the handle nothing happens. 


But when a user with a recognized face (indicated by a green LED) tries to turn the handle, the lock is disengaged allowing the latch to move.



With just a few lines of code and some external hardware, we were able to create a fully-functioning face-controlled e-lock that works entirely offline just by using the SLN-VIZN-IOT. Not to mention the fact that there was no need for any ML experience whatsoever.

Because the SLN-VIZN-IOT was designed with flexibility in mind, all sorts of use cases can be supported with only minimal effort when compared to a face recognition implemented from scratch. By using the production-ready software that comes provided with the kit, it’s now possible to add local (no cloud connectivity necessary) face and emotion recognition capabilities to all sorts of products in record time.

We hope this guide was helpful in showing you how to jumpstart your face recognition project with the power of the SLN-VIZN-IOT. 

Labels (1)
Version history
Last update:
‎05-12-2020 05:51 PM
Updated by: