Відео

What version did you download

I'm not sure. It's quite an old tutorial and the libraries have been updated several times. Are you using 1.0.5?

Not for a while. I have some in progress but nothing completed.

I have no idea what that is. What did you want to do?

@@robotzeroone I want the ESP32-CAM connected to a TFT screen to be able to make all that connection to the server and from there do the facial recognition. Without using a laptop or computer to address the web server.

@@robotzeroone I don't know if I explain myself.

@@brianalmeida529 You could combine this project and this one: robotzero.one/esp32-cam-scene-analyzer/ to send the image to a server and print the name of the recognised face back into the screen.

I don't know what that error is. If that's the standard camerawebserver example it sounds maybe a hardware problem.

I don't think there is any problem with the camera module. Because "wifi scan", "LED blinking switch webserver" and other programs were running fine.

error is "No Core Dump Partition Found"

@@aminulhaque3253 Looking on Reddit it seems the camera is broken. Did you try the camerawebserver example? It's an old tutorial but here robotzero.one/esp32-cam-arduino-ide/ with 1.0.4 of the ESP32 hardware libraries. If it doesn't work I think your camera is broken.

I just realized that I'm going to be missing pins to place my fingerprint sensor. This is a double authentication project that I want to do, incorporating fingerprint sensors and more facial recognition.

@@brianalmeida529 The problem has always been the pins. So many are used by the camera that there are very few available. I think my project will still work with the correct pins set.

@@robotzeroone What do you think of the wrover cam, will it be useful for my project? I have the following components: relay, fingerprint, that's all. And if there are more pins left over I want to use one or two buttons.

@@brianalmeida529 I'm not sure. If you can find a pin diagram you can see which pins are available and what interfaces they offer. This project was mine with most pins used so might give some help robotzero.one/esp32-cam-scene-analyzer/

@@robotzeroone Thanks, this will help me. Then I'll show you how it turned out more or less.

I think it shows that the first time it runs because it has to make the file.

Which version of the ESP32 hardware library do you have installed. 1.0.4 or 1.0.5 works best.

I don't really have to time to extract it but the code is in this sketch: github.com/robotzero1/esp32cam-selfiecam/blob/master/SelfieCam.ino

HI, I think it will work with those versions of the IDE but the ESP Hardware library has to be be 1.0.4 or 5.

yep i managed to do that but i cant see the partition scheme for some reason@@robotzeroone

Can you change the board to AIThinker ESP32 Cam? Maybe that will work with Huge App selected

when i select ESP32 CAM it wont let me change the partition scheme at all@@robotzeroone

Shouldn't be too hard to remove what you don't need from the code. The hardest part is knowing which pins to connect. robotzero.one/esp32-cam-tft-selfie/

@@robotzeroone thanks for the response

@@robotzeroone sadly it didnt help much because i cant properly remove the parts of code i dont need they just turn into errors

Is this the example that comes with the IDE? Are you on version 1.0.4 or 5 of the hardware libraries? I know the recent version only does face detection but not recognition.

Probably. Does it have the same trigger voltage? They have to be quite sensitive because the ESP doesn't output much current.

This library file - /cef_image.h I don't recognise that.

@@robotzeroone me too. Your not using this library in the code! What can I do to fix it? Does it have to do with the esp32 filesystem uploader? Because I dont need the numbers and the esp32 filesystem uploader is not working, I dont see the tool once I go to arduino.

@@justadude5160 I'm not sure. The project is quite old and the libraries change. Did you see it on my blog? robotzero.one/esp32-cam-tft-selfie/ You could try from scratch and use the libraries that have the same date as the publish date of the tutorial.

​@@robotzerooneI managed to get it working by changing the esp32 library to an older version, thanks

Here's a version that doesn't need a permanent connection. robotzero.one/esp32-face-door-entry/ I think there's also a version in the comments of this tutorial. Other tutorials that might be helpful: robotzero.one/esp32-camera-projects/

Thank you so much for your reply. Will try and get back to you. Take care

You would have to know android development but I imagine you can connect to the IP address within an app

It will be way too slow for that. It can barely manage face recognition and in the latest libraries they've dropped that.

@@robotzeroone Thank you for the answer... I understand from this that I cannot use Esp32 because it has limited capabilities, right?

Yep. @@muhammadsalem9924

On the photo or the screen? Screen will be the wrong screen chosen in the settings. Photo will be a dodgy camera connection or broken camera.

@@robotzeroone photo. Looks like you wearing the night vision goggle. But it is normal when the cam in video mode. So stranger.

Yep. Here's a version without the webserver: robotzero.one/esp32-face-door-entry/ You can mix and match from various tutorials on my blog.

Yep. It's just a demo to show the circuit.

All here: github.com/robotzero1/esp32cam-selfiecam

Which version of the ESP32 Hardware Library are you using?

long interval = 5000; // open lock for ... milliseconds

for which part? There's a wiring diagram on my blog,

Does the CameraWebServer example work for you? Could be a dodgy board.

@@robotzeroone Its not an issue with the board bc when ran on a version like 2.0.8 i get no issues and smooth video. I am working on this as a final project and need it complete by next week. if you could share a way i could get the lock to open based on at least face detection and would most llikley have to be on 2.0.8 because of the video issue. Please get back to me when you can thanks.

to specify yes i ran the camerawebserver example on 2.0.8 and no issues.

if there is somewhere better we can communicate that would be great

@@gabesantiago6941 I don't really have time to look at this but you just need to find the code that does the face detection in 2.0.8 and hook in some code to open the relay. I don't know why the camera would be OK in 2.0.8 and not 1.0.5. It's not a Wi-fi thing is is it? Here's two tutorials that use 1.0.4 for face detection. Might be something helpful in them: github.com/robotzero1/esp32cam-selfiecam/blob/master/SelfieCam.ino robotzero.one/unlock-door-face-detection-sonoff/

probably the wrong version of the ESP32 Hardware library is installed. Try 1.0.5

@@robotzeroone how can i get this version ? i am searching that and i cant find it sorry if my english its so bad

You can comment on this project in my blog and I might be able to help if it's something quick.

I can't really help because I've never used Platformio. This will only work with the version 1.0.4 and 1.0.5 of the ESP32 hardware libraries so I would start there.

@@robotzerooneOK, thanks. But why didn't you import the files fr_forward.c and fr_flash.c into the project? You just imported their .h files.Can you explain it to me a little bit if you can.

@@user-waldner I think that it just tells the script to use those files in the Expressif libraries.

@@robotzeroone Hello sir, I have a new question again, dl_matrix3du_t *image_matrix = dl_matrix3du_alloc(1, 320, 240, 3); Why are you allocating a huge array? Because when I am doing functional operations, sometimes the webpage will get stuck, and then the board will restart the server. Because when I am doing functional operations, sometimes the webpage will get stuck, and then the board will restart the server. The error prompted by the serial port is related to the allocation of the matrix.

@@user-waldner Explained here: techtutorialsx.com/2020/06/13/esp32-camera-face-detection/

Maybe with the new versions of the ESP32 hardware libraries the HTML is encoded differently.

This is the kit. You can see a good photo of the motor. Probably more info on the Elegoo site as well. amzn.to/2yJ3OAG

Probably the wrong version of the ESP32 hardware library. Try 1.0.5

or can i put an mic output to the website? using the arduino

@@michaelguablas1669 I don't really understand. What is it you want to do?

Which version of the ESP32 hardware library do you have installed?

@@robotzeroone 1.0.6

@@that_one_who_knows Not sure if it will work. Try 1.0.5 with the github files.

HI, It should work. Could be the power supply or connectors aren't providing enough current. The ESP32 uses a lot when the Wi-Fi starts up. Does the CameraWebServer example work? .

@@robotzeroone Yes, the CameraWebServer example does work, but only when connected to the computer via shield. As for the current, I checked that there was current, so it's probably just because i need an FTDI (which is confusing, but i gotta try something.) I just got an FTDI programmer, so i can also include a simple circuit to test while connected to my computer. I just need to see if my theory works. If it doesn't, then idk what i'm doing wrong...

@@robotzeroone Besides the hardware though, I noticed that when using the webserver for this code, it does not stop sampling the faces, and the count goes and repeats 1, 2, 3, 4, 7 instead of doing only the four samples 1, 2, 3, 4 and then stopping and saving the face like in your video. Any idea why this is happening? is there a part of the code i need to change? Edit: I realized that the states defined ENROLL_COMPLETE and DELETE_ALL are not used at all. Is there a reason why? Edit2: Does it matter if i have the AI thinker or the Wrover kit Module? Because I have the AI thinker, and it only works if i put the board setting that way in tools, but the partition scheme selection disappears, so i can't select the face recognition. Edit3: Okay, i was able to fix the face capturing after seeing someone else go through the same problem, so I fixed the face count variable if statement to stop counting when less-than-or-equal-to 0 since idk how to make the variable not skip 0.

@@KristellaJackson IT sounds like somethings not set up correctly from the beginning. I had the face counting above 4 problem once when testing this on a different versions and it was because I'd missed a step. I don't remember exactly. Are you following this tutorial? robotzero.one/access-control-with-face-recognition/ ? Using the files here: github.com/robotzero1/esp32cam-access-control it should just work.

@@robotzeroone Okay, yeah, i did not realize i only had to put the partitions.csv file in the same folder with the sketch file. Fixing the code back to test for fr_flash.h return value is 0, it works properly now; the enrolled faces stay permanently in the webserver even when powering off and then back on the ESP32-CAM. Thank you 👍

Probably the wrong ESP32 hardware libraries. Check the github for help