Simple ESP32-CAM Object Detection

Train an ESP32-CAM for Object Detection using a free Edge Impulse account. All without writing any code!
Article with full instructions: dronebotworkshop.com/esp32-ob...
More articles and tutorials: dronebotworkshop.com
Join the conversation on the forum: forum.dronebotworkshop.com
Subscribe to the newsletter and stay in touch: dronebotworkshop.com/subscribe/
Object Detection is a Machine Learning (ML) technique that uses a computer with a camera to identify classes of objects it has been trained to recognize. Although this used to require massive computers and large datasets, it is now possible to perform decent object detection using microcontrollers.
While most microcontroller-based object detection projects use 64-bit chips like the powerful Kendryte K210, you can also use a 32-bit ESP32 to get the job done. And that’s exactly what we will do today - take a 9-dollar ESP32-CAM board and train it to detect objects.
I’ll be using a lantern battery and Robie, a 40-year-old Radio Shack robot, as my two “test objects” We’ll train a model using the powerful online neural network services of Edge Impulse; you’ll need an account, but don’t worry, it’s completely free.
Once we have our model trained, we’ll export it in the form of an Arduino Library. The library even includes a sample sketch we can use to test our model; no code to write! I’ll try it on an ESP32-CAM and ESP-EYE board.
Here is the Table of Contents for today's video:
00:00 - Introduction
02:34 - Object Detection
06:27 - Edge Impulse
08:15 - Workflow
10:47 - Look at ESP32-CAM & ESP-EYE
12:48 - Capturing Images - Webcam Setup
14:34 - Edge Impulse Setup
15:39 - Image Capture
17:57 - Label Images
20:32 - Create an Impulse
24:52 - Export to an Arduino Library
25:25 - Importing Library to Arduino IDE
27:41 - Testing with ESP-EYE
29:43 - Testing with ESP32-CAM
32:04 - EloquentESP32CAM Library
34:14 - Collect Images Sketch
35:35 - Capturing Images - ESP32-CAM
39:12 - Import & Label with Edge Impulse
42:46 - Build & Deploy Impulse
46:45 - Testing with ESP32-CAM
48:28 - EloquentArduino Code & Demo
51:32 - Conclusion
53:25 - Robie the Robot!
I will actually show you two methods of capturing images of your subject(s) - using a webcam or using the ESP32-CAM board itself. I’ll also introduce you to a library that can make doing all of this a lot easier.
The results are good, good enough to consider using the ESP32-CAM as an inexpensive object detection sensor.
Hope you enjoy the video!

[@Dronebotworkshop]

Hi everyone - It's come to my attention that there has been someone masquerading as myself, responding to some comments here with a link to a Telegram chat to win a prize from me. THIS IS A SCAM, I am not holding a contest, nor do I have a Telegram account. PLEASE DO NOT RESPOND TO THESE MESSAGES!! It's happening on a lot of my videos, I'm taking steps to remove them manually, but as I have 162 videos, it will take some time. If you do run across a suspicious comment, I would appreciate you letting me know at [e-mail]. Thanks! Bill (The real one!)

[@GabbyShimoni]

I've just taken some training on the topic of deep learning and this video completes my quest for an easy way to incorporate the subject into my students projects. Thanks a lot for another well done video

[@dakcn]

As usual, another fantastic tutorial. I really like having both the video and the written articles. Thank you!!

[@anthonyd4539]

I really appreciate the time you take to give instructions. Your lessons give me the confidence to try new things and peak over the barriers that I will soon knock down.

[@ShawnHymel]

This is fantastic! You do a great job at clearly explaining the steps. And gotta love a Robie cameo :)

[@raulsw9088]

👍🏻

[@edgarbaylosis1523]

Finally the video that i've been lookin for months! Thanks!

[@svlaststar]

Wow!!! Excellent video and project. IT is one that's been on my list for a while and you just inspired me to move it up the list. Thanks for making your videos. Great job!!!

[@Gengh13]

As someone without any knowledge in this area this was a nice introduction.

[@mirekcolakov1730]

And of course, as always, thank you and well done. Thorough, detailed and explained so well!!!

[@hamradio3716]

Excellent. A lot of prep work. Thank you.😊

[@davydbar]

Excellent tutorial!! Like all your videos, I want to thank you, since you have inspired me to continue in robotics projects. Greetings from Santa Fe, Argentina.

[@LIMUNQUE]

Thanks a lot, I was spending time trying to use opencv for an academical project and I didn't think to use this.

[@rogermathura]

Glad i subscribed, so i didn't miss your excellent tutorial . Thanks.

[@fostee1]

As always just what I was looking for. Best tutorials thank you

[@donkeypoopdragon]

Awesome topic, Bill! Keep 'em comin!

[@adolphtrudeau]

This is excellent, DroneBot Workshop--I can't wait to tell my kid and his robotics teammates

[@christopherleadholm6677]

This is *just* the type of video I was hoping for!

[@mrcairo4]

It is a very easy training scenario. I've tested with voice recognition and now I'll try it with different "generic" humans vs. dogs and cats.

[@gerdsfargen6687]

This is very sexy time for an esp32 cam. Cant wait to try this project out. Its brilliant!

[@fremsoft]

Thank you for your explanation,I think I will introduce Edge Impulse in my classroom! This is very inspiring video 😄👍🏻

[@ahmedihab9754]

I was searching for such a video for days

[@dldldl8028]

Great tutorial, exactly what I'm looking for. Thanks a lot

[@aladinmovies]

You are nice teacher. Love your understanding material.

[@mikedoyle7546]

Great and precise video tutorial.

[@TheSelfUnemployed]

I just built a tiny robot using an esp32-cam and now im going to attempt to add this badass functionality. great video and ive always loved your channel. thanks for sharing your knowledge with us!

[@karimaaqa]

hello when using edge impulse, which board did you chose? the espressif esp32 eye? since i am going to get the esp32 cam (the one without usb)

[@micmike]

Nice demo of capabilities. Thanks

[@michaelbarakat6095]

Great effort has been done in this tutorial thank you very much for this amazing work ☺️

[@NZHippie]

Thank you Bill, great video...

[@jeffschroeder4805]

Another informative video - Thank You

[@Bianchi77]

Well done, keep it up, thanks for sharing it with us:)

[@TheOldKid]

A tutorial that is greatly needed is on the s3 wroom cam board.

[@terrificmenace]

Very informative 👏🏻 thank you 😊

[@manusaini2192]

A great contents, i love your video's. Lot's of love ❤ From India 🇮🇳. Thanks for sharing knowledge with us.

[@devluz]

Oh love seeing another one with the ESP32-Cam! I already have two of these in my network as cheap surveillance camera running my own software. I am curious if there are any good upgrade paths for higher frame rate / higher resolution cameras? Also more ports available to control a robot would be neat.

[@mirekcolakov1730]

use espnow or mqtt protocol so multiple esp32s can talk to each other and do the heavy compute on a proper machine that is connected wireless so the robot does not need to carry it...

[@itamariliuk]

Amazing. Thank you so much.

[@rhandynastor4866]

good as always

[@ej21lalitkumawat81]

Lots of love and great video Proffecer love you From India❤❤

[@UTube4075]

Aap engineer ho? Working professional??

[@Jean-Naimar]

Great tutorial 👍

[@rubialugattimoreira1978]

Perfect! Sharing knologe always.

[@Dazza_Doo]

You know these are the types projects I'm looking for (I've found the DroidsVader will be pleased), this can be extremely useful. Thank you!

[@andrebullitt7212]

Dude, you are the coolest!

[@brendanowen7563]

Great video. Looking forward to having a go with my ESP32 Cam. My only comment is you probably got the 100% on your second example because you had multiple images that are nearly exactly the same. These would have been randomly distributed between the 80% training and 20% testing sets. You probably ended up with same images used in training and testing which would not give you a good measure of confidence. Again another great video, thank you.

[@Ducerobot]

I like this module . Ai camera works well. It's on the base for my project. 😁👍

[@threewheeler624]

Thank you for a good video. Have a great week. I miss Radio Shacks in the malls.

[@barnacode]

Great video, very helpful. Can I implement the object detection on the car robot you built? Has it been done? Any help is appreciated.

[@AungZawMoe-zh8fu]

What a great job sir. This vd help me a lot for my project. And what a cute robot ROBIE

[@conversationwithgod272]

Great job from great man.

[@ulzs7683]

This ones not in the ESP32 playlist is it? I think its a good addition to the collection along with the 2024 guide.

[@yuriyyatsyshyn9717]

Thanks. Very usefull!!

[@Abdul-bv4ku]

That's great! What was the frame rate achieved?

[@coolprojects5297]

Great Job Thank you very much

[@arthus1971]

Bravooo!!! A powerfull Like to you!!! Thank you for teach us

[@mdrakibhossen8681]

helo sir, this is a great video. i have one question. can i add a servo to this project so that when it detects robot or battery the servo will rotate?

[@sudedemmanuel2975]

Thank you very useful

[@optipwr40]

Do you know if it's possible to use RectLabel (An offline image annotation labeling tool) which has limited export options? For custom object detection I'd prefer not to upload all of my images into an online source.

[@keshav2136]

I've a ESP-32. I'll try it.

[@luiklaus3024]

hi, is there a limit on how many different objects the esp32 cam can recognize?

[@digihz_data]

Just great!

[@jamespletcher7574]

Clever outro!

[@ibekwekingsley8538]

Is there a way to get the live feed video of the image classified in a bounding box?

[@catalinalb1722]

Hi, I want to do a number plate recognition for my gate. Would this be good enough?

[@MUFATECHSOLUTIONS]

thank you sir

[@nonel7215]

Thanks!

[@Dronebotworkshop]

And thank you as well!

[@abrarsahil2476]

Can we add a kaggle data set for training for obstacle detection in Edge impulses?

[@vinothn4228]

Better way of explaining ❗

[@renznormanpalma6312]

Might be new here, but I would like to ask if its possible to control other peripheral based from the label it detected?

[@calmrelaxed7626]

does the esp32 cam needs to be connected to com port always in order to work or is it wireless?

[@raulsw9088]

Hi, you have very good projects. Can this object detection project be done with the Xiao esp32s3 sense?

[@transientaardvark6231]

Does this have any way to train using the COCO data set ?

[@w74anheda]

hi, thanks for your helpful tutorial, i have a question: how can connect and stream cam a esp32 cam with cable to pc or raspberry pi??

[@user-hx5tg7yz2y]

you should make a turntable that turns at an extremely slow rate for taking images.

[@mirekcolakov1730]

Hello, what is the upper limit of the amount of objects the library and EspCam can handle ??? thank you

[@varun8154]

can you use esp32 cam to control like servo based on object detection

[@alexandrecrt]

Master, what if the object is on a mat? Does detection also happen assertively? Thank you very much.

[@azizghazni2605]

is this suitable for eyes closure detection?

[@angelaponte4413]

Maravilloso video, si les sale error al subir a la tarjeta Esp32, revisen la versión de la tarjeta que están usando, a mí me sirvió con la versión 2.0.10

[@karaniii]

I have been trying to do some research on this but haven't found answers yet. Can edge impulse be used for face recognition?

[@darshanpatel805]

@Dronebotworkshop I like all your content can you please make video of speech recognition using ESP? Thank you

[@letgodigital6787]

Hello sir Which variable from the codes store's the output of the detection I want to make decisions wit h the Arduino to turn on a fan

[@erment55]

I follow your educational videos with great interest.I would be very happy if you could share a video on how to use the esp32 camera module in IoT Blynk 2.0.

[@rostykromanyshyn6616]

great thanks a lot

[@christopherleadholm6677]

I totally remember that coin eating robot thing, btw. The good ol' days...

[@opdino8577]

i just have one question can the detection be with voice. like can the model speak what kind of object is in front

[@ibekwekingsley8538]

Is there a way to view the live feed with the object detected in a bounding box

[@saysanvicente5238]

Is it possible to make an alarm connected to esp32?For example, the program object to detect is paper and plastic only. When the camera detected the plastic it will make an alarm. We're planning to make a detection system of paper and plastic waste for our capstone project.

[@iwandp8698]

very helpfull

[@stephankempa6914]

A slow running turn table and the ESP auto capturing.

[@cybercrazy1059]

How many same object can esp32-cam detect at the same time?

[@sam-el]

How can object detection be implemented using de1-soc? I will like a guide please

[@AliHassan-jw5yq]

Hi, can you please guide me how we use esp32 cam for face detection and object detection in a single project??

[@vivekent4422]

hey the eye board is not available i have an esp 32 its base and arduino is there any way i can still use edge impulse as these boards are not mentioned in the list of the software

[@AtlantaTerry]

Is there a parts list somewhere?

[@HEILZEUS]

can i like, upload my own trained model onto edge impulse ? also, please mention that the auto labeller is enabled in your project because you have an enterprise edition and not the free community edition

[@OverlordYashSrivastav]

i am using an ov7670 camera ,can it work with esp32 or pi pico

[@marsupialsrock010]

Is this what Captcha is for? Training?

[@wisnual-rasyid2598]

hmmm, i install the eloquent library but cant upload, it say the library is missing :(

[@LucasFerreira-qz1ib]

Hello, i've tried running this code on the arduino IDE, using the esp32 cam, but the code won't compile, perhaps it's a configuration issue in the ide?

[@Junior-fq8jo]

Hi Mr Bill thx for ur helpful video. I followed all the steps of your process but during that i have met a problem as :"error for the ardûino borad Ai thinker Esp32-cam. If u can help me thx😊

[@stanwilliams7337]

Hi, I just found your site and have a question. I am working on automating my roll off roof observatory and have one big issue. I need a way to let the software I use know that my telescope is in it's "Parked" or "Home" position so the roof can open or close safely. I was thinking an ultrasonic or an IR sensor. Does one of these sound reasonable, or do you recommend something else? I am using an arduino board with sensors to say when the roof is open or closed. I am using NINA, RRCI and SkyRoof to operate a garage door opener to move the roof. Any suggestions would be greatly appreciated.

[@stanwilliams7337]

I received an email from you saying I won a prize. Is this from you or is it a phishing scam?

[@carlosporras6905]

Señor lo amooo