November 17, 2019 by Shakhizat Nurgaliyev
The M5Stick-V is an AIOT (AI + IOT) camera with machine vision capabilities.
| 1 | M5StickV | ||
|
|
|||
The M5StickV is a tiny AI + IOT camera almost the size in diameter of 2 coins, the price is around $27.00 which for some could seem expensive for such a tiny camera but it does pack some decent specs. The camera is powered by incredibly powerful machine learning AI chip Kendryte K210, an edge computing system-on-chip with dual-core 64bit RISC-V CPU and state-of-art neural network processor, which makes it perfect and ready-to-use for:
The M5StickV comes in a nice package containing the M5StickV itself and USB-A to USB-C cable.
Hardware specification can be found on the package.
A more detailed specification can be found in official website of M5Stack.
There are 3 methods to burn firmware for Windows OS:
Using EasyLoader tool
Using Kflash GUI
Using command prompt
pip3 install kflashAfter finishing installation, run the following command
kflash.exe -p COM3 M5StickV_Firmware_1022_beta.kfpkgIt will print:
[32m[1m[INFO][0m COM Port Selected Manually: COM3 [0m
[32m[1m[INFO][0m Default baudrate is 115200 , later it may be changed to the value you set. [0m
[32m[1m[INFO][0m Trying to Enter the ISP Mode... [0m
._
[32m[1m[INFO][0m Automatically detected goE/kd233 [0m
[32m[1m[INFO][0m Greeting Message Detected, Start Downloading ISP [0m
Downloading ISP: |=====================================================================================| 100.0% 8kiB/s
[32m[1m[INFO][0m Booting From 0x80000000 [0m
[32m[1m[INFO][0m Wait For 0.1 second for ISP to Boot [0m
[32m[1m[INFO][0m Boot to Flashmode Successfully [0m
[32m[1m[INFO][0m Selected Flash: On-Board [0m
[32m[1m[INFO][0m Initialization flash Successfully [0m
[32m[1m[INFO][0m Extracting KFPKG ... [0m
[32m[1m[INFO][0m Writing maixpy.bin into 0x00000000 [0m
Programming BIN: |=====================================================================================| 100.0% 9kiB/s
[32m[1m[INFO][0m Writing m5stickv_resources.img into 0x00d00000 [0m
Programming BIN: |=====================================================================================| 100.0% 10kiB/s
[32m[1m[INFO][0m Writing facedetect.kmodel into 0x00300000 [0m
Programming BIN: |=====================================================================================| 100.0% 9kiB/s
[32m[1m[INFO][0m Rebooting... [0msudo pip3 install kflashUsing Kflash burn firmware image
sudo kflash -b 1500000 -B goE M5StickV_Firmware_1022_beta.kfpkgYou will see the similar following output as above.
Open terminal and run the following command
sudo pip3 install kflashIf you receive an error after installation, try the following command:
sudo python -m pip install kflash
sudo python3 -m pip install kflash
sudo pip install kflash
sudo pip2 install kflashEnter the following command
sudo kflash -b 1500000 -B goE M5StickV_Firmware_1022_beta.kfpkgYou must see the same output as above on Linux section.
For MacOS and Linux:
sudo apt-get install screenUsing screen utility connect to M5stickV via serial communication
sudo screen /dev/ttyUSB0 115200It will print:
[MAIXPY]Pll0:freq:832000000
[MAIXPY]Pll1:freq:398666666
[MAIXPY]Pll2:freq:45066666
[MAIXPY]cpu:freq:416000000
[MAIXPY]kpu:freq:398666666
[MAIXPY]Flash:0xc8:0x17
open second core...
gc heap=0x80215060-0x80295060
[MaixPy] init end
__ __ _____ __ __ _____ __ __
| \/ | /\ |_ _| \ \ / / | __ \ \ \ / /
| \ / | / \ | | \ V / | |__) | \ \_/ /
| |\/| | / /\ \ | | > < | ___/ \ /
| | | | / ____ \ _| |_ / . \ | | | |
|_| |_| /_/ \_\ |_____| /_/ \_\ |_| |_|
M5StickV by M5Stack : https://m5stack.com/
M5StickV Wiki : https://docs.m5stack.com
Co-op by Sipeed : https://www.sipeed.com
[MAIXPY]: result = 0
[MAIXPY]: numchannels = 1
[MAIXPY]: samplerate = 44100
[MAIXPY]: byterate = 88200
[MAIXPY]: blockalign = 2
[MAIXPY]: bitspersample = 16
[MAIXPY]: datasize = 158760
init i2c2
[MAIXPY]: find ov7740When connected, it will automatically enter Maixpy UI. Now the device is running the default program code, you can terminate it by Ctrl+C.
For Windows
Enter the following commands in your terminal of MacOS and Linux. For Windows use PuTTY.
import lcd
lcd.init()
lcd.draw_string(100, 100, "hello world", lcd.RED, lcd.BLACK)
The MaixPY IDE is a cloned version of the OpenMV IDE. M5StickV supports the OpenMV and MicroPython development environments.
By default face detection model and program code was already preinstalled. Here's how it works.
import sensor
import image
import KPU as kpu
sensor.reset()
sensor.set_pixformat(sensor.RGB565)
sensor.set_framesize(sensor.QVGA)
sensor.run(1)
task = kpu.load(0x300000)
anchor = (1.889, 2.5245, 2.9465, 3.94056, 3.99987, 5.3658, 5.155437, 6.92275, 6.718375, 9.01025)
a = kpu.init_yolo2(task, 0.5, 0.3, 5, anchor)
while(True):
img = sensor.snapshot()
code = kpu.run_yolo2(task, img)
if code:
for i in code:
print(i)
a = img.draw_rectangle(i.rect())
a = kpu.deinit(task)Press the Run button, and the board displays live video from the camera to the MaixPy IDE.
The accuracy is pretty good considering we are running it on a $27 board. This is truly impressive and revolutionary.
This board is not ideal though, it lacks analog inputs, microphone, WiFi, and Bluetooth. However, it is a great camera with AI capabilities that can be used for face recognition, object or shape detection and many other detection activities. Also, this is an awesome dev kit to get started with the Kendryte K210 RISC-V core.
I hope you found this guide useful and thanks for reading. If you have any questions or feedback? Leave a comment below. Stay tuned!
