Hey everybody! Welcome back to the AI on the Edge series!

In Lesson 49, we’re doing something that’s just plain cool — we’re turning that tiny little SSD1306 OLED into a live animated avatar that follows your face and hands in real time!

That’s right — no more static images, and no more being tethered to your desktop and monitor. This time your OLED is alive! Move your head, smile, raise your eyebrows, wave your hands… and the little display instantly mirrors your movements like a miniature digital twin. It’s like having a tiny version of yourself living on that 128×64 screen!

What You’ll Learn in This Lesson:

How to use MediaPipe Face Mesh to track 468 facial landmarks
How to combine it with Hand Tracking at the same time
The trick to projecting those landmarks onto a tiny OLED in real time
How to draw both points and connecting lines to create smooth, animated avatars
How to keep everything running fast on the Raspberry Pi

This is one of those projects that makes people stop and say “Wait… how is that even possible on that little screen?!”

By the end of this lesson, you’ll have a foundation for building all kinds of fun real-time avatar projects — everything from simple face mirrors to gesture-controlled characters and beyond.

Why This Is Awesome:

We’re not just displaying video on the big screen anymore — we’re compressing all that AI power down onto a cheap little OLED. This is real AI on the Edge stuff, and it looks incredibly impressive for how simple the hardware is.

Whether you want to build interactive displays, robot faces, wearable tech, or just blow your friends’ minds, this lesson gives you the core technique to make it happen.

Ready to make your OLED come alive?

Fire up the code, run it, and watch your tiny digital self start moving with you. Then start experimenting! Try drawing just the eyes and mouth, or just the hands, or maybe even turn it into a little stick figure avatar.

As always, I want to see what you create with this! Drop your versions in the comments or tag me — I love seeing the creative stuff you guys come up with. Don’t just copy and paste my code. Show you really understand the video by making the code yout own.

Enough talk, lets get this party started!

# ====================================================================
# DISCLAIMER:
# This code is provided as-is for educational and experimental 
# purposes only. The author makes no representations or warranties of 
# any kind concerning the safety, suitability, or accuracy of this 
# code. Use at your own risk. The author assumes no liability for any 
# damages, system failures, security breaches, or network issues 
# resulting from the use or implementation of this script.
# ====================================================================

import cv2
import time
from picamera2 import Picamera2
import mediapipe as mp
import numpy as np
import board
import adafruit_ssd1306
from PIL import Image, ImageDraw, ImageFont
import time

wOLED = 128
hOLED = 64
I2C_ADDRESS = 0x3C
i2c = board.I2C()
oled = adafruit_ssd1306.SSD1306_I2C(wOLED, hOLED, i2c, addr = I2C_ADDRESS)

W=1280
H=720
tStart = time.time()
fps = 0

piCam = Picamera2(1)
piCam.preview_configuration.main.size = (W, H)
piCam.preview_configuration.main.format = "RGB888"
piCam.preview_configuration.controls.FrameRate = 60
piCam.preview_configuration.align()
piCam.configure("preview")
piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))
fontFace = cv2.FONT_HERSHEY_SIMPLEX
fontThickness = int(W/425)
fontScale = H*.0015
fontColor = (0,0,255)

faceMesh = mp.solutions.face_mesh.FaceMesh(
    
    max_num_faces=1,
    refine_landmarks=True,
    min_detection_confidence=0.5,
    min_tracking_confidence=0.5)

hands = mp.solutions.hands.Hands(
    model_complexity = 0,
    min_detection_confidence = .5,
    min_tracking_confidence = .5,
    max_num_hands=2 )

cv2.namedWindow('Camera',cv2.WINDOW_GUI_NORMAL)
cv2.moveWindow('Camera',0,65)
cv2.resizeWindow('Camera',W,H)
while True:
    deltaT = time.time() - tStart
    tStart=time.time()
    fps = fps*.95 + (1/deltaT)*.05
    frame = piCam.capture_array()
    frame = cv2.flip(frame, -1)

    rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    faceResults = faceMesh.process(rgb)
    handResults = hands.process(rgb)
    
    image = Image.new("1",(wOLED,hOLED))
    draw = ImageDraw.Draw(image)
    
    if handResults.multi_hand_landmarks:
        for myHands in handResults.multi_hand_landmarks:
            landMarks = myHands.landmark
            lmAll =[]
            for lm in landMarks:
                x = int(lm.x*W)
                y = int(lm.y*H)
                cv2.circle(frame,(x,y),10,(255,0,0),-1)
                xOLED = int(lm.x*wOLED)
                yOLED = int(lm.y*hOLED)
                draw.point((xOLED,yOLED), fill =255)
                lmAll.append((xOLED,yOLED))
            #Uncomment out these lines if you want
            #Lines between the finger joints
#             for idx in mp.solutions.hands.HAND_CONNECTIONS:
#                 lineStart = idx[0]
#                 lineEnd = idx[1]
#                 pt1 = lmAll[lineStart]
#                 pt2 = lmAll[lineEnd]
#                 draw.line((pt1[0],pt1[1],pt2[0],pt2[1]),width=1,fill=255)
#                              
    if faceResults.multi_face_landmarks:
        for faceLandMarks in faceResults.multi_face_landmarks:
            #print(faceLandmarks)
            lmAll = []
            for lm in faceLandMarks.landmark:
                xOLED = int(lm.x*wOLED)
                yOLED = int(lm.y*hOLED)
                draw.point((xOLED,yOLED), fill = 255)
                #print(xOLED,yOLED)
                lmAll.append((xOLED,yOLED))
            for idx in mp.solutions.face_mesh.FACEMESH_CONTOURS:
                #print(idx)
                lineStart = idx[0]
                lineEnd = idx[1]
                pt1 = lmAll[lineStart]
                pt2 = lmAll[lineEnd]
                draw.line((pt1[0],pt1[1],pt2[0],pt2[1]), width=1, fill = 255)



    oled.image(image)
    oled.show()


    myText = "FPS: "+str(round(fps,1))
    cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)

    cv2.imshow("Camera", frame)

    if cv2.waitKey(1) == ord('q'):
        break

cv2.destroyAllWindows()
piCam.stop()
oled.fill(0)
oled.show()

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

# ====================================================================

# DISCLAIMER:

# This code is provided as-is for educational and experimental

# purposes only. The author makes no representations or warranties of

# any kind concerning the safety, suitability, or accuracy of this

# code. Use at your own risk. The author assumes no liability for any

# damages, system failures, security breaches, or network issues

# resulting from the use or implementation of this script.

# ====================================================================

import cv2

import time

from picamera2 import Picamera2

import mediapipe as mp

import numpy as np

import board

import adafruit_ssd1306

from PIL import Image, ImageDraw, ImageFont

import time

wOLED = 128

hOLED = 64

I2C_ADDRESS = 0x3C

i2c = board.I2C()

oled = adafruit_ssd1306.SSD1306_I2C(wOLED, hOLED, i2c, addr = I2C_ADDRESS)

W=1280

H=720

tStart = time.time()

fps = 0

piCam = Picamera2(1)

piCam.preview_configuration.main.size = (W, H)

piCam.preview_configuration.main.format = "RGB888"

piCam.preview_configuration.controls.FrameRate = 60

piCam.preview_configuration.align()

piCam.configure("preview")

piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))

fontFace = cv2.FONT_HERSHEY_SIMPLEX

fontThickness = int(W/425)

fontScale = H*.0015

fontColor = (0,0,255)

faceMesh = mp.solutions.face_mesh.FaceMesh(

max_num_faces=1,

refine_landmarks=True,

min_detection_confidence=0.5,

min_tracking_confidence=0.5)

hands = mp.solutions.hands.Hands(

model_complexity = 0,

min_detection_confidence = .5,

min_tracking_confidence = .5,

max_num_hands=2 )

cv2.namedWindow('Camera',cv2.WINDOW_GUI_NORMAL)

cv2.moveWindow('Camera',0,65)

cv2.resizeWindow('Camera',W,H)

while True:

deltaT = time.time() - tStart

tStart=time.time()

fps = fps*.95 + (1/deltaT)*.05

frame = piCam.capture_array()

frame = cv2.flip(frame, -1)

rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

faceResults = faceMesh.process(rgb)

handResults = hands.process(rgb)

image = Image.new("1",(wOLED,hOLED))

draw = ImageDraw.Draw(image)

if handResults.multi_hand_landmarks:

for myHands in handResults.multi_hand_landmarks:

landMarks = myHands.landmark

lmAll =[]

for lm in landMarks:

x = int(lm.x*W)

y = int(lm.y*H)

cv2.circle(frame,(x,y),10,(255,0,0),-1)

xOLED = int(lm.x*wOLED)

yOLED = int(lm.y*hOLED)

draw.point((xOLED,yOLED), fill =255)

lmAll.append((xOLED,yOLED))

#Uncomment out these lines if you want

#Lines between the finger joints

# for idx in mp.solutions.hands.HAND_CONNECTIONS:

# lineStart = idx[0]

# lineEnd = idx[1]

# pt1 = lmAll[lineStart]

# pt2 = lmAll[lineEnd]

# draw.line((pt1[0],pt1[1],pt2[0],pt2[1]),width=1,fill=255)

if faceResults.multi_face_landmarks:

for faceLandMarks in faceResults.multi_face_landmarks:

#print(faceLandmarks)

lmAll = []

for lm in faceLandMarks.landmark:

xOLED = int(lm.x*wOLED)

yOLED = int(lm.y*hOLED)

draw.point((xOLED,yOLED), fill = 255)

#print(xOLED,yOLED)

lmAll.append((xOLED,yOLED))

for idx in mp.solutions.face_mesh.FACEMESH_CONTOURS:

#print(idx)

lineStart = idx[0]

lineEnd = idx[1]

pt1 = lmAll[lineStart]

pt2 = lmAll[lineEnd]

draw.line((pt1[0],pt1[1],pt2[0],pt2[1]), width=1, fill = 255)

oled.image(image)

oled.show()

myText = "FPS: "+str(round(fps,1))

cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)

cv2.imshow("Camera", frame)

if cv2.waitKey(1) == ord('q'):

break

cv2.destroyAllWindows()

piCam.stop()

oled.fill(0)

oled.show()

AI On the Edge, Raspberry Pi

AI on the Edge LESSON 48: Hand Detection in OpenCV and MediaPipe on the Raspberry Pi

September 6, 2026 admin

In Lesson 48 we’re taking a huge step forward — we’re bringing real-time hand detection to the Raspberry Pi using the incredible power of MediaPipe combined with OpenCV and the Pi Camera. This is one of those projects that feels like pure magic when you see it working: the camera picks up your hands instantly, tracks all 21 landmarks on each hand, draws beautiful connections in real time, and does it all right on the edge with no cloud required.

You’ll learn how to set up MediaPipe’s Hands solution for reliable multi-hand tracking, how to efficiently process frames from the Picamera2 library, flip and convert images for proper display, and draw professional-looking landmarks and connections with custom styling. We also keep a smooth FPS counter running so you can see exactly how well your Pi is performing.

This lesson is exciting because hand tracking is the foundation for so many advanced gesture-control projects — think touchless interfaces, sign language recognition, robotic control, virtual instruments, and interactive art installations. Once you have solid hand detection running smoothly on your Raspberry Pi, the creative possibilities are almost endless.

So grab your Raspberry Pi, fire up the camera, and let’s get those hands dancing on the screen! As always, the full code is waiting for you below. Watch the video, follow along, and then start experimenting — I can’t wait to see what you build with this!

Let’s make something awesome!

# ====================================================================
# DISCLAIMER:
# This code is provided as-is for educational and experimental 
# purposes only. The author makes no representations or warranties of 
# any kind concerning the safety, suitability, or accuracy of this 
# code. Use at your own risk. The author assumes no liability for any 
# damages, system failures, security breaches, or network issues 
# resulting from the use or implementation of this script.
# ====================================================================

import cv2
import time
from picamera2 import Picamera2
import mediapipe as mp
piCam = Picamera2(1)
W=1280
H=720
tStart = time.time()
fps = 0
RES = (W,H)
piCam.preview_configuration.main.size = RES
piCam.preview_configuration.main.format = "RGB888"
piCam.preview_configuration.controls.FrameRate=60
piCam.preview_configuration.align()
piCam.configure("preview")
piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))
fontFace = cv2.FONT_HERSHEY_SIMPLEX
fontThickness = int(W/425)
fontScale = H*.0015
fontColor = (0,0,255)

hands = mp.solutions.hands.Hands(
    model_complexity =  0,
    min_detection_confidence = .5,
    min_tracking_confidence = .5,
    max_num_hands =2)
mpDraw = mp.solutions.drawing_utils

handStyle = mpDraw.DrawingSpec(
    color = (100, 0, 255),
    thickness = 6,
    circle_radius = 8)
connectionStyle = mpDraw.DrawingSpec(
    color = (0,255,0),
    thickness = 5)

while True:
    deltaT = time.time() - tStart
    tStart=time.time()
    fps = fps*.95 + (1/deltaT)*.05
    frame= piCam.capture_array()
    frame=cv2.flip(frame,-1)
    rgbFrame= cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    results = hands.process(rgbFrame)
    if results.multi_hand_landmarks:
        for handLM in results.multi_hand_landmarks:
            mpDraw.draw_landmarks(
                frame,
                handLM,
                mp.solutions.hands.HAND_CONNECTIONS,
                handStyle,
                connectionStyle)
        
    myText = "FPS: "+str(round(fps,1))
    cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)
    cv2.imshow("Camera", frame)
    cv2.moveWindow("Camera",0,60)
    if cv2.waitKey(1)==ord('q'):
        break
cv2.destroyAllWindows()
print('Program Terminated')

# ====================================================================

# DISCLAIMER:

# This code is provided as-is for educational and experimental

# purposes only. The author makes no representations or warranties of

# any kind concerning the safety, suitability, or accuracy of this

# code. Use at your own risk. The author assumes no liability for any

# damages, system failures, security breaches, or network issues

# resulting from the use or implementation of this script.

# ====================================================================

import cv2

import time

from picamera2 import Picamera2

import mediapipe as mp

piCam = Picamera2(1)

W=1280

H=720

tStart = time.time()

fps = 0

RES = (W,H)

piCam.preview_configuration.main.size = RES

piCam.preview_configuration.main.format = "RGB888"

piCam.preview_configuration.controls.FrameRate=60

piCam.preview_configuration.align()

piCam.configure("preview")

piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))

fontFace = cv2.FONT_HERSHEY_SIMPLEX

fontThickness = int(W/425)

fontScale = H*.0015

fontColor = (0,0,255)

hands = mp.solutions.hands.Hands(

model_complexity = 0,

min_detection_confidence = .5,

min_tracking_confidence = .5,

max_num_hands =2)

mpDraw = mp.solutions.drawing_utils

handStyle = mpDraw.DrawingSpec(

color = (100, 0, 255),

thickness = 6,

circle_radius = 8)

connectionStyle = mpDraw.DrawingSpec(

color = (0,255,0),

thickness = 5)

while True:

deltaT = time.time() - tStart

tStart=time.time()

fps = fps*.95 + (1/deltaT)*.05

frame= piCam.capture_array()

frame=cv2.flip(frame,-1)

rgbFrame= cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

results = hands.process(rgbFrame)

if results.multi_hand_landmarks:

for handLM in results.multi_hand_landmarks:

mpDraw.draw_landmarks(

frame,

handLM,

mp.solutions.hands.HAND_CONNECTIONS,

handStyle,

connectionStyle)

myText = "FPS: "+str(round(fps,1))

cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)

cv2.imshow("Camera", frame)

cv2.moveWindow("Camera",0,60)

if cv2.waitKey(1)==ord('q'):

break

cv2.destroyAllWindows()

print('Program Terminated')

AI On the Edge, Raspberry Pi

AI on the Edge LESSON 47: Emotion Detector Using MediaPipe, OpenCV and Raspberry Pi

September 3, 2026 admin

In this exciting lesson, we build a real-time Emotion Detector that can recognize basic human emotions using just a Raspberry Pi, a camera, and the power of MediaPipe. The system watches your face and identifies whether you look Neutral, Smiling, Surprised, or Angry — and then reacts instantly by changing the color of a NeoPixel ring and displaying the detected emotion on both the camera preview and a small OLED display.Using MediaPipe’s Face Mesh solution, the program tracks 468 facial landmarks in real time. From these points, we carefully calculate key facial ratios — such as eye openness, mouth width, and mouth height relative to head width. These measurements allow us to create simple but effective rules that distinguish between different emotional expressions. For example, a wide mouth combined with raised cheeks indicates a smile, while a small mouth opening and narrowed eyes suggests anger.The project beautifully integrates several powerful technologies:

Picamera2 for high-performance camera capture
OpenCV for image processing and on-screen text display
MediaPipe Face Mesh for accurate facial landmark detection
NeoPixel RGB ring that lights up in different colors depending on the detected emotion
SSD1306 OLED display that shows both the emotion name and a simplified wireframe of your face

One of the most satisfying parts of this project is seeing the NeoPixel ring instantly change color to match your emotion — green for happy, cyan for surprised, red for angry, and yellow for neutral. The OLED also mirrors the detected emotion, making the entire system feel alive and responsive.This lesson is a fantastic step forward in understanding how to combine computer vision with physical outputs on the edge. You will learn how to extract meaningful measurements from facial landmarks, build rule-based emotion logic, and synchronize visual feedback across multiple devices (camera preview, NeoPixel, and OLED).By the end of Lesson 47, you will have a working real-time emotion recognition system running entirely on a Raspberry Pi — a great foundation for more advanced projects like mood-reactive lights, interactive robots, or even assistive technology.

# ====================================================================
# DISCLAIMER:
# This code is provided as-is for educational and experimental 
# purposes only. The author makes no representations or warranties of 
# any kind concerning the safety, suitability, or accuracy of this 
# code. Use at your own risk. The author assumes no liability for any 
# damages, system failures, security breaches, or network issues 
# resulting from the use or implementation of this script.
# ====================================================================

import cv2
import time
from picamera2 import Picamera2
import mediapipe as mp
import numpy as np
#################################
import board
import adafruit_ssd1306
from PIL import Image, ImageDraw, ImageFont
import time

import neopixel_spi as neopixel
import colorsys
spi=board.SPI()
LC = 12
PO = neopixel.GRB
strip = neopixel.NeoPixel_SPI(spi, LC,
            pixel_order=PO,auto_write=False)
strip.fill(0)
strip.show()

wOLED = 128
hOLED = 64
I2C_ADDRESS = 0x3C
i2c = board.I2C()
oled = adafruit_ssd1306.SSD1306_I2C(wOLED, hOLED, i2c, addr = I2C_ADDRESS)
fontLarge = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf",14)
#################################
W=1280
H=720
tStart = time.time()
fps = 0

piCam = Picamera2(1)
piCam.preview_configuration.main.size = (W, H)
piCam.preview_configuration.main.format = "RGB888"
piCam.preview_configuration.controls.FrameRate = 60
piCam.preview_configuration.align()
piCam.configure("preview")
piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))
fontFace = cv2.FONT_HERSHEY_SIMPLEX
fontThickness = int(W/425)
fontScale = H*.0015
fontColor = (0,0,255)

faceMesh = mp.solutions.face_mesh.FaceMesh(
    
    max_num_faces=1,
    refine_landmarks=True,
    min_detection_confidence=0.5,
    min_tracking_confidence=0.5)

cv2.namedWindow('Camera',cv2.WINDOW_GUI_NORMAL)
cv2.moveWindow('Camera',0,65)
cv2.resizeWindow('Camera',W,H)
while True:
    deltaT = time.time() - tStart
    tStart=time.time()
    fps = fps*.95 + (1/deltaT)*.05
    frame = piCam.capture_array()
    frame = cv2.flip(frame, -1)

    rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    results = faceMesh.process(rgb)
    
    if results.multi_face_landmarks:
        for faceLandMarks in results.multi_face_landmarks:
            #print(faceLandmarks)
            image = Image.new("1",(wOLED,hOLED))
            draw = ImageDraw.Draw(image)
            lmAll = []
            j=0
            for lm in faceLandMarks.landmark:
                xOLED = int(lm.x*wOLED)
                yOLED = int(lm.y*hOLED)
                draw.point((xOLED,yOLED), fill = 255)
                #print(xOLED,yOLED)
                lmAll.append((xOLED,yOLED))
                x = int(lm.x*W)
                y = int(lm.y*H)
                #cv2.putText(frame,str(j),(x,y),fontFace, .3, (0,255,0),1)
                j = j+1
            for idx in mp.solutions.face_mesh.FACEMESH_CONTOURS:
                #print(idx)
                lineStart = idx[0]
                lineEnd = idx[1]
                pt1 = lmAll[lineStart]
                pt2 = lmAll[lineEnd]
                draw.line((pt1[0],pt1[1],pt2[0],pt2[1]), width=1, fill = 255)

            keyLM = faceLandMarks.landmark
            
            headRight = 323
            headLeft = 93
            headWidth = keyLM[headRight].x -keyLM[headLeft].x
            cv2.putText(frame,str(headRight),(int(keyLM[headRight].x*W),int(keyLM[headRight].y*H)),fontFace, .3, (0,255,0),1)
            cv2.putText(frame,str(headLeft),(int(keyLM[headLeft].x*W),int(keyLM[headLeft].y*H)),fontFace, .3, (0,255,0),1)
            
            leftEyeTop = 159
            leftEyeBottom = 145
            leftEyeHeight =keyLM[leftEyeBottom].y -keyLM[leftEyeTop].y
            leftEyeLeft = 130
            leftEyeRight = 154
            leftEyeWidth = keyLM[leftEyeRight].x -keyLM[leftEyeLeft].x
            leftEyeRatio = leftEyeHeight/leftEyeWidth*100 
            
            rightEyeTop = 385
            rightEyeBottom = 253
            rightEyeHeight = keyLM[rightEyeBottom].y -keyLM[rightEyeTop].y
            rightEyeLeft = 465
            rightEyeRight = 249
            rightEyeWidth = keyLM[rightEyeRight].x -keyLM[rightEyeLeft].x
            rightEyeRatio = rightEyeHeight/rightEyeWidth*100
            
            eyeRatio = (rightEyeRatio +leftEyeRatio)/2
            
            mouthLeft = 61
            mouthRight = 291
            mouthTop = 13
            mouthBottom = 15
            mouthWidth = keyLM[mouthRight].x-keyLM[mouthLeft].x
            mouthHeight = keyLM[mouthBottom].y-keyLM[mouthTop].y
            mouthWidthRatio = mouthWidth/headWidth *100
            mouthHeightRatio = mouthHeight/headWidth*100
            
            print(round(mouthWidthRatio,1),round(mouthHeightRatio,1),round(eyeRatio,1))
            color = (255,255,0)
            emotion = 'NEUTRAL'
            if mouthWidthRatio >37 and mouthHeightRatio>4 and eyeRatio<74:
                emotion = 'SMILE'
                color = (0,255,0)
            if mouthWidthRatio <37 and mouthHeightRatio>10 and eyeRatio>70:
                emotion = 'SURPRISED'
                color = (0,255,255)
            if mouthWidthRatio <32 and mouthHeightRatio<4 and eyeRatio<57:
                emotion = 'ANGRY'
                color = (0,0,255)               
            cv2.putText(frame, emotion, (int(W*.1),int(H*.15)),
                    fontFace, 2.2, color, 4)
            strip.fill((color[2],color[1],color[0]))
            strip.show()
            #print('XXXXXXXXXXXXXXXXXXx')
            draw.text((0,0),emotion,font=fontLarge,fill=255)
            oled.image(image)
            oled.show()
        #print()

#     if results.multi_face_landmarks:
#         for faceLandmarks in results.multi_face_landmarks:
#             # Optional second pass for glow effect
#             mp.solutions.drawing_utils.draw_landmarks(
#                 image=frame,
#                 landmark_list=faceLandmarks,
#                 connections=mp.solutions.face_mesh.FACEMESH_TESSELATION,
#                 landmark_drawing_spec=None,
#                 connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(
#                     color=(100, 100, 100),
#                     thickness=1
#                 )
#             )
#             # === Best looking translucent mesh ===
#             mp.solutions.drawing_utils.draw_landmarks(
#                 image=frame,
#                 landmark_list=faceLandmarks,
#                 connections=mp.solutions.face_mesh.FACEMESH_CONTOURS,
#                 landmark_drawing_spec=None,                    # Hide dots
#                 connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(
#                     color=(0, 0, 180),      # Bright cyan/teal
#                     thickness=7
#                 )
#             )
# 
#             # Optional: Extra detail on irises (makes eyes look better)
#             mp.solutions.drawing_utils.draw_landmarks(
#                 image=frame,
#                 landmark_list=faceLandmarks,
#                 connections=mp.solutions.face_mesh.FACEMESH_IRISES,
#                 landmark_drawing_spec=None,
#                 connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(
#                     color=(255, 0, 0),
#                     thickness=7
#                 )
#             )

    myText = "FPS: "+str(round(fps,1))
    cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)

    cv2.imshow("Camera", frame)

    if cv2.waitKey(1) == ord('q'):
        break

cv2.destroyAllWindows()
piCam.stop()
oled.fill(0)
oled.show()
strip.fill(0)
strip.show()

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

# ====================================================================

# DISCLAIMER:

# This code is provided as-is for educational and experimental

# purposes only. The author makes no representations or warranties of

# any kind concerning the safety, suitability, or accuracy of this

# code. Use at your own risk. The author assumes no liability for any

# damages, system failures, security breaches, or network issues

# resulting from the use or implementation of this script.

# ====================================================================

import cv2

import time

from picamera2 import Picamera2

import mediapipe as mp

import numpy as np

#################################

import board

import adafruit_ssd1306

from PIL import Image, ImageDraw, ImageFont

import time

import neopixel_spi as neopixel

import colorsys

spi=board.SPI()

LC = 12

PO = neopixel.GRB

strip = neopixel.NeoPixel_SPI(spi, LC,

pixel_order=PO,auto_write=False)

strip.fill(0)

strip.show()

wOLED = 128

hOLED = 64

I2C_ADDRESS = 0x3C

i2c = board.I2C()

oled = adafruit_ssd1306.SSD1306_I2C(wOLED, hOLED, i2c, addr = I2C_ADDRESS)

fontLarge = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf",14)

#################################

W=1280

H=720

tStart = time.time()

fps = 0

piCam = Picamera2(1)

piCam.preview_configuration.main.size = (W, H)

piCam.preview_configuration.main.format = "RGB888"

piCam.preview_configuration.controls.FrameRate = 60

piCam.preview_configuration.align()

piCam.configure("preview")

piCam.start()

textLowerLeft = (int(W*.01),int(H*.05))

fontFace = cv2.FONT_HERSHEY_SIMPLEX

fontThickness = int(W/425)

fontScale = H*.0015

fontColor = (0,0,255)

faceMesh = mp.solutions.face_mesh.FaceMesh(

max_num_faces=1,

refine_landmarks=True,

min_detection_confidence=0.5,

min_tracking_confidence=0.5)

cv2.namedWindow('Camera',cv2.WINDOW_GUI_NORMAL)

cv2.moveWindow('Camera',0,65)

cv2.resizeWindow('Camera',W,H)

while True:

deltaT = time.time() - tStart

tStart=time.time()

fps = fps*.95 + (1/deltaT)*.05

frame = piCam.capture_array()

frame = cv2.flip(frame, -1)

rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

results = faceMesh.process(rgb)

if results.multi_face_landmarks:

for faceLandMarks in results.multi_face_landmarks:

#print(faceLandmarks)

image = Image.new("1",(wOLED,hOLED))

draw = ImageDraw.Draw(image)

lmAll = []

j=0

for lm in faceLandMarks.landmark:

xOLED = int(lm.x*wOLED)

yOLED = int(lm.y*hOLED)

draw.point((xOLED,yOLED), fill = 255)

#print(xOLED,yOLED)

lmAll.append((xOLED,yOLED))

x = int(lm.x*W)

y = int(lm.y*H)

#cv2.putText(frame,str(j),(x,y),fontFace, .3, (0,255,0),1)

j = j+1

for idx in mp.solutions.face_mesh.FACEMESH_CONTOURS:

#print(idx)

lineStart = idx[0]

lineEnd = idx[1]

pt1 = lmAll[lineStart]

pt2 = lmAll[lineEnd]

draw.line((pt1[0],pt1[1],pt2[0],pt2[1]), width=1, fill = 255)

keyLM = faceLandMarks.landmark

headRight = 323

headLeft = 93

headWidth = keyLM[headRight].x -keyLM[headLeft].x

cv2.putText(frame,str(headRight),(int(keyLM[headRight].x*W),int(keyLM[headRight].y*H)),fontFace, .3, (0,255,0),1)

cv2.putText(frame,str(headLeft),(int(keyLM[headLeft].x*W),int(keyLM[headLeft].y*H)),fontFace, .3, (0,255,0),1)

leftEyeTop = 159

leftEyeBottom = 145

leftEyeHeight =keyLM[leftEyeBottom].y -keyLM[leftEyeTop].y

leftEyeLeft = 130

leftEyeRight = 154

leftEyeWidth = keyLM[leftEyeRight].x -keyLM[leftEyeLeft].x

leftEyeRatio = leftEyeHeight/leftEyeWidth*100

rightEyeTop = 385

rightEyeBottom = 253

rightEyeHeight = keyLM[rightEyeBottom].y -keyLM[rightEyeTop].y

rightEyeLeft = 465

rightEyeRight = 249

rightEyeWidth = keyLM[rightEyeRight].x -keyLM[rightEyeLeft].x

rightEyeRatio = rightEyeHeight/rightEyeWidth*100

eyeRatio = (rightEyeRatio +leftEyeRatio)/2

mouthLeft = 61

mouthRight = 291

mouthTop = 13

mouthBottom = 15

mouthWidth = keyLM[mouthRight].x-keyLM[mouthLeft].x

mouthHeight = keyLM[mouthBottom].y-keyLM[mouthTop].y

mouthWidthRatio = mouthWidth/headWidth *100

mouthHeightRatio = mouthHeight/headWidth*100

print(round(mouthWidthRatio,1),round(mouthHeightRatio,1),round(eyeRatio,1))

color = (255,255,0)

emotion = 'NEUTRAL'

if mouthWidthRatio >37 and mouthHeightRatio>4 and eyeRatio<74:

emotion = 'SMILE'

color = (0,255,0)

if mouthWidthRatio <37 and mouthHeightRatio>10 and eyeRatio>70:

emotion = 'SURPRISED'

color = (0,255,255)

if mouthWidthRatio <32 and mouthHeightRatio<4 and eyeRatio<57:

emotion = 'ANGRY'

color = (0,0,255)

cv2.putText(frame, emotion, (int(W*.1),int(H*.15)),

fontFace, 2.2, color, 4)

strip.fill((color[2],color[1],color[0]))

strip.show()

#print('XXXXXXXXXXXXXXXXXXx')

draw.text((0,0),emotion,font=fontLarge,fill=255)

oled.image(image)

oled.show()

#print()

# if results.multi_face_landmarks:

# for faceLandmarks in results.multi_face_landmarks:

# # Optional second pass for glow effect

# mp.solutions.drawing_utils.draw_landmarks(

# image=frame,

# landmark_list=faceLandmarks,

# connections=mp.solutions.face_mesh.FACEMESH_TESSELATION,

# landmark_drawing_spec=None,

# connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(

# color=(100, 100, 100),

# thickness=1

# )

# # === Best looking translucent mesh ===

# mp.solutions.drawing_utils.draw_landmarks(

# image=frame,

# landmark_list=faceLandmarks,

# connections=mp.solutions.face_mesh.FACEMESH_CONTOURS,

# landmark_drawing_spec=None, # Hide dots

# connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(

# color=(0, 0, 180), # Bright cyan/teal

# thickness=7

# )

# # Optional: Extra detail on irises (makes eyes look better)

# mp.solutions.drawing_utils.draw_landmarks(

# image=frame,

# landmark_list=faceLandmarks,

# connections=mp.solutions.face_mesh.FACEMESH_IRISES,

# landmark_drawing_spec=None,

# connection_drawing_spec=mp.solutions.drawing_utils.DrawingSpec(

# color=(255, 0, 0),

# thickness=7

# )

myText = "FPS: "+str(round(fps,1))

cv2.putText(frame,myText,textLowerLeft,fontFace,fontScale,fontColor,fontThickness)

cv2.imshow("Camera", frame)

if cv2.waitKey(1) == ord('q'):

break

cv2.destroyAllWindows()

piCam.stop()

oled.fill(0)

oled.show()

strip.fill(0)

strip.show()

These are the schematics we are using for todays project. If you are taking the class, you should already have these components connected:

Fusion Hat Circuit Diagram — This is the circuit we will use moving forward in the class

SSD1306 OLED Connected to the Fusion AI Hat

AI On the Edge, Raspberry Pi

AI on the Edge LESSON 46: Ultimate Dazzleing Running Rainbow On a NeoPixel Ring

August 30, 2026 admin

In this lesson we create one of the most beautiful and satisfying NeoPixel effects — a smooth, continuous, running rainbow on a 12-LED ring. We affectionately have named this pattern the “Runbow”. This is the “ultimate” version because it is both visually stunning and highly educational, as we explore two fundamentally different programming approaches the Runbow.

The first method is the most intuitive: we place a fixed rainbow across the ring (each LED gets a different hue equally spaced around the color wheel) and then simply rotate the entire pattern one position at a time. This approach is easy to understand because it mimics physically moving a colorful wheel. We save the color of the first pixel, shift all the other pixels one place, and then place the saved color at the end. It feels like a conveyor belt of color circling around the ring.

The second method is more elegant and mathematically pure. Instead of storing and shifting colors, we recalculate the color of every LED on every frame using a moving offset. For each LED we compute its hue as (i / LED_COUNT + offset) % 1.0, where i is the LED’s position and offset is a value that slowly increases over time. This creates a perfectly smooth rainbow that flows around the ring without ever shifting raw RGB values. Because we regenerate the pattern fresh each cycle, there is no risk of color corruption or accumulated errors.

Both techniques produce a gorgeous running rainbow, but they teach very different programming mindsets. The first method helps you deeply understand array manipulation and data movement. The second method introduces the powerful concept of using mathematics and offsets to create motion — a technique used frequently in advanced LED animations, games, and visual effects. In the end, the offset method tends to look smoother and is easier to extend with additional effects (such as brightness pulsing), but both approaches are valuable skills for any embedded AI or IoT developer working with addressable LEDs.You can adjust the speed of the rainbow by changing how much you increment the offset each loop or by modifying the delay. Once you master these two methods, you will be able to create almost any animated pattern you can imagine on your NeoPixel ring. Lets get this party started!

Here is the code we developed in this video:

# ====================================================================
# DISCLAIMER:
# This code is provided as-is for educational and experimental 
# purposes only. The author makes no representations or warranties of 
# any kind concerning the safety, suitability, or accuracy of this 
# code. Use at your own risk. The author assumes no liability for any 
# damages, system failures, security breaches, or network issues 
# resulting from the use or implementation of this script.
# ====================================================================
 
import time
import board
import neopixel_spi as neopixel
import colorsys
spi = board.SPI()
LC = 12
PO = neopixel.GRB
strip = neopixel.NeoPixel_SPI(spi,
            LC,pixel_order=PO,
            auto_write=False)
strip.fill(0)
strip.show()
def hsv2rgb(h):
    r, g, b = colorsys.hsv_to_rgb(h,1,1)
    return ( int(255*r), int(255*g),int(255*b))
ts=1
ts2=.1
ts3=.05
try:
    for i in range(12):
        hue = i/12
        color = hsv2rgb(hue)
        strip[i] = color
    strip.show()
    time.sleep(ts)
    for lap in range(70):
        startColor = strip[0]
        for i in range(11):
            strip[i]= strip[i+1]
        strip[11] = startColor
        time.sleep(ts2)
        strip.show()
    offset = 0
    while True:
        for i in range(12):
            hue = (i/12 + offset)%1
            strip[i] = hsv2rgb(hue)
        strip.show()
        offset = (offset + 1/100)%1
        time.sleep(ts3)
    
except KeyboardInterrupt:
    strip.fill(0)
    strip.show()
    print("That's All Folks!")

# ====================================================================

# DISCLAIMER:

# This code is provided as-is for educational and experimental

# purposes only. The author makes no representations or warranties of

# any kind concerning the safety, suitability, or accuracy of this

# code. Use at your own risk. The author assumes no liability for any

# damages, system failures, security breaches, or network issues

# resulting from the use or implementation of this script.

# ====================================================================

import time

import board

import neopixel_spi as neopixel

import colorsys

spi = board.SPI()

LC = 12

PO = neopixel.GRB

strip = neopixel.NeoPixel_SPI(spi,

LC,pixel_order=PO,

auto_write=False)

strip.fill(0)

strip.show()

def hsv2rgb(h):

r, g, b = colorsys.hsv_to_rgb(h,1,1)

return ( int(255*r), int(255*g),int(255*b))

ts=1

ts2=.1

ts3=.05

try:

for i in range(12):

hue = i/12

color = hsv2rgb(hue)

strip[i] = color

strip.show()

time.sleep(ts)

for lap in range(70):

startColor = strip[0]

for i in range(11):

strip[i]= strip[i+1]

strip[11] = startColor

time.sleep(ts2)

strip.show()

offset = 0

while True:

for i in range(12):

hue = (i/12 + offset)%1

strip[i] = hsv2rgb(hue)

strip.show()

offset = (offset + 1/100)%1

time.sleep(ts3)

except KeyboardInterrupt:

strip.fill(0)

strip.show()

print("That's All Folks!")

This is the schematic we are using to connect the NeoPixel ring:

AI On the Edge, Raspberry Pi

AI on the Edge LESSON 45: Adding a NeoPixel Ring To Your Raspberry Pi Project

August 26, 2026 admin

In Lesson 45 of our AI on the Edge series, we take our Fusion AI Lab Kit to the next level by adding a 12-pixel NeoPixel ring. This lesson bridges the gap between pure AI processing and vibrant physical output, showing how your edge AI projects can communicate visually with the real world in a beautiful and engaging way.

We begin by setting up the NeoPixel ring using the SPI interface on the SunFounder Fusion Hat. After importing the necessary libraries (time, board, neopixel_spi, colorsys, and math), we create a simple and reusable hsv2rgb() function that converts Hue values (0.0 to 1.0) into RGB colors that the NeoPixels can understand. This function becomes the foundation for smooth rainbow effects later in the lesson.The lesson starts with basic pixel control. We manually light up each of the 12 pixels one at a time using different colors (red, green, blue, cyan, magenta, yellow, etc.). This slow, deliberate approach lets you clearly see how individual pixel addressing works and helps students understand the coordinate system of the ring.Next, we explore full-ring control by making the entire ring blink between bright red and blue. We then move into motion with a running green pixel moving across a red background — a great introduction to animation techniques. This is followed by a more advanced chasing effect where a blue pixel chases a green pixel around a dim red background.One of the highlights of this lesson is the gentle pulsating aqua effect. Using a sine wave (math.sin), we create a smooth breathing/pulsing animation where the brightness of the color rises and falls naturally. This technique produces a very professional and visually pleasing result that students can easily adapt for future projects.

Finally, we create a rainbow effect. First, we display a uniform rainbow where all 12 pixels show the same color that cycles smoothly through the entire spectrum.

We finish the lesson by assigning the homework. The homework is for you to create the classic running rainbow (what I like to call a “Runbow”), where the colors flow continuously around the ring — one of the most popular and impressive NeoPixel animations.

This lesson reinforces important programming concepts including loops, functions, color theory (HSV vs RGB), timing control, and animation techniques, while giving students an exciting visual payoff. The skills learned here open the door to creating stunning visual feedback for future AI projects — whether it’s status indicators, emotional displays, or attention-grabbing outputs from your edge AI models.

# ====================================================================
# DISCLAIMER:
# This code is provided as-is for educational and experimental 
# purposes only. The author makes no representations or warranties of 
# any kind concerning the safety, suitability, or accuracy of this 
# code. Use at your own risk. The author assumes no liability for any 
# damages, system failures, security breaches, or network issues 
# resulting from the use or implementation of this script.
# ====================================================================
 
import time
import board
import neopixel_spi as neopixel
import colorsys
import math

def hsv2rgb(h):
    r, g, b = colorsys.hsv_to_rgb(h, 1, 1)
    return (int(r*255),int(g*255),int(b*255))

spi = board.SPI()
LED_COUNT = 12
PO = neopixel.GRB
strip = neopixel.NeoPixel_SPI(spi,
        LED_COUNT,pixel_order=PO,
        auto_write = False)
strip.fill(0)
strip.show()

strip[0] = (255,0,0)
strip.show()
ts=.1
time.sleep(ts)

strip[1] = (0,255,0)
strip.show()
time.sleep(ts)

strip[2] = (0,0,255)
strip.show()
time.sleep(.1)

strip[3] = (0,255,255)
strip.show()
time.sleep(ts)

strip[4] = (255,0,255)
strip.show()
time.sleep(.1)

strip[5] = (255,255,0)
strip.show()
time.sleep(.1)

strip[6] = (255,100,0)
strip.show()
time.sleep(ts)

strip[7] = (128,0,255)
strip.show()
time.sleep(.1)

strip[8] = (0,255,128)
strip.show()
time.sleep(.1)

strip[9] = (255,0,128)
strip.show()
time.sleep(ts)

strip[10] = (128,255,0)
strip.show()
time.sleep(ts)

strip[11] = (0,128,255)
strip.show()
time.sleep(.1)
ts2=5
for i in range(5):
    strip.fill((255,0,0))
    strip.show()

    time.sleep(ts2)
    
    strip.fill((0,0,255))
    strip.show()
    time.sleep(ts2)
    
strip.fill((255,0,0))
strip.show()
ts3=1
time.sleep(ts3)
ts4=.3
for lap in range(3):
    for i in range(12):
        strip[i] = (0,255,0)
        strip.show()
        time.sleep(ts4)
        strip[i] = (255,0,0)

strip.fill((40,0,0))
strip.show()
time.sleep(ts3)

for lap in range(4):
    for i in range(12):
        greenPos = i
        bluePos = (i + 6)%12
        
        strip[greenPos] = (0,255,0)
        strip[bluePos] = (0,0,255)
        strip.show()
        time.sleep(ts4)
        
        strip[greenPos] = (40,0,0)
        strip[bluePos] = (40,0,0)
        
for laps in range(6):
    for step in range(360):
        brightness= (math.sin(step*3.1415/180)+1)/2
        r = 0
        g = int(255*brightness)
        b = int(50*brightness)
        strip.fill((r,g,b))
        strip.show()
        time.sleep(.005)
        
for laps in range(5):
    for step in range(100):
        hue = step/100
        color = hsv2rgb(hue)
        strip.fill(color)
        strip.show()
        time.sleep(ts)


strip.fill(0)
strip.show()

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

# ====================================================================

# DISCLAIMER:

# This code is provided as-is for educational and experimental

# purposes only. The author makes no representations or warranties of

# any kind concerning the safety, suitability, or accuracy of this

# code. Use at your own risk. The author assumes no liability for any

# damages, system failures, security breaches, or network issues

# resulting from the use or implementation of this script.

# ====================================================================

import time

import board

import neopixel_spi as neopixel

import colorsys

import math

def hsv2rgb(h):

r, g, b = colorsys.hsv_to_rgb(h, 1, 1)

return (int(r*255),int(g*255),int(b*255))

spi = board.SPI()

LED_COUNT = 12

PO = neopixel.GRB

strip = neopixel.NeoPixel_SPI(spi,

LED_COUNT,pixel_order=PO,

auto_write = False)

strip.fill(0)

strip.show()

strip[0] = (255,0,0)

strip.show()

ts=.1

time.sleep(ts)

strip[1] = (0,255,0)

strip.show()

time.sleep(ts)

strip[2] = (0,0,255)

strip.show()

time.sleep(.1)

strip[3] = (0,255,255)

strip.show()

time.sleep(ts)

strip[4] = (255,0,255)

strip.show()

time.sleep(.1)

strip[5] = (255,255,0)

strip.show()

time.sleep(.1)

strip[6] = (255,100,0)

strip.show()

time.sleep(ts)

strip[7] = (128,0,255)

strip.show()

time.sleep(.1)

strip[8] = (0,255,128)

strip.show()

time.sleep(.1)

strip[9] = (255,0,128)

strip.show()

time.sleep(ts)

strip[10] = (128,255,0)

strip.show()

time.sleep(ts)

strip[11] = (0,128,255)

strip.show()

time.sleep(.1)

ts2=5

for i in range(5):

strip.fill((255,0,0))

strip.show()

time.sleep(ts2)

strip.fill((0,0,255))

strip.show()

time.sleep(ts2)

strip.fill((255,0,0))

strip.show()

ts3=1

time.sleep(ts3)

ts4=.3

for lap in range(3):

for i in range(12):

strip[i] = (0,255,0)

strip.show()

time.sleep(ts4)

strip[i] = (255,0,0)

strip.fill((40,0,0))

strip.show()

time.sleep(ts3)

for lap in range(4):

for i in range(12):

greenPos = i

bluePos = (i + 6)%12

strip[greenPos] = (0,255,0)

strip[bluePos] = (0,0,255)

strip.show()

time.sleep(ts4)

strip[greenPos] = (40,0,0)

strip[bluePos] = (40,0,0)

for laps in range(6):

for step in range(360):

brightness= (math.sin(step*3.1415/180)+1)/2

r = 0

g = int(255*brightness)

b = int(50*brightness)

strip.fill((r,g,b))

strip.show()

time.sleep(.005)

for laps in range(5):

for step in range(100):

hue = step/100

color = hsv2rgb(hue)

strip.fill(color)

strip.show()

time.sleep(ts)

strip.fill(0)

strip.show()

In this class we are using this as our standard components. You should already have the core circuit built and should already have the OLED connected. Today you will add the NeoPixel array.

Our core circuit is:

Last week we also added the SSD1306 OLED Display.

And finally today we add the NeoPixel ring from the Fusion AI Lab Kit.

Technology Tutorials

AI on the Edge LESSON 49: Live Animated Avatars on the SSD1306 OLED With OpenCV

What You’ll Learn in This Lesson:

Why This Is Awesome:

AI on the Edge LESSON 48: Hand Detection in OpenCV and MediaPipe on the Raspberry Pi

AI on the Edge LESSON 47: Emotion Detector Using MediaPipe, OpenCV and Raspberry Pi

AI on the Edge LESSON 46: Ultimate Dazzleing Running Rainbow On a NeoPixel Ring

AI on the Edge LESSON 45: Adding a NeoPixel Ring To Your Raspberry Pi Project

Making The World a Better Place One High Tech Project at a Time. Enjoy!