Core Electronics Forum

Raspberry Pi Tensor Flow Lite Output

I’m currently working on a security camera using Open CV and Tensorflow lite on my Raspberry Pi 3b+. I’m having trouble rendering a output from Tensorflow Lite to a GPIO pin (5V). My ideal goal is when it detects a person it will run a continuous output to the 5V GPIO pin so that i can use a breadboard for a custom circuit.

Ive tried some code that ive found but it hasnt worked as of yet.

Cheers!

Good morning Rorque,

Welcome to the forum! :partying_face:

Ooh how interesting, have you got a link to a tutorial that you’ve followed to set this up? Also, what does the script that you’re using to trigger the GPIO look like? Are you able to throw it onto the forum here? I’m curious to see how it interfaces with tensorflow when objects (i.e. people :sweat_smile:) are detected to find out what’s going on with it.

1 Like

Yeah, ill get the URL for the install tutorial;

How To Run TensorFlow Lite on Raspberry Pi for Object Detection - YouTube (video guide)

TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi/Raspberry_Pi_Guide.md at master · EdjeElectronics/TensorFlow-Lite-Object-Detection-on-Android-and-Raspberry-Pi · GitHub (written guide)

Its using a object detection from google (can be found within the tutorial) and the detection script comes with the set up (TFLite_detection_webcam.py)
i cant upload it because its a .py file

Hey Rorque,

No worries, if you can copy-paste it between three ` on the top and bottom of the script that should autoformat to code highlighting for python. Although obviously if it’s on the Pi and too difficult to move over to your machine your on at the moment to post it I completely understand :slightly_smiling_face: just want to make sure that we’re looking at the same script before I start making recommendations for changes that could cause errors.

######## Webcam Object Detection Using Tensorflow-trained Classifier #########
#
# Author: Evan Juras
# Date: 10/27/19
# Description: 
# This program uses a TensorFlow Lite model to perform object detection on a live webcam
# feed. It draws boxes and scores around the objects of interest in each frame from the
# webcam. To improve FPS, the webcam object runs in a separate thread from the main program.
# This script will work with either a Picamera or regular USB webcam.
#
# This code is based off the TensorFlow Lite image classification example at:
# https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/examples/python/label_image.py
#
# I added my own method of drawing boxes and labels using OpenCV.

# Import packages
import os
import argparse
import cv2
import numpy as np
import sys
import time
from threading import Thread
import importlib.util

# Define VideoStream class to handle streaming of video from webcam in separate processing thread
# Source - Adrian Rosebrock, PyImageSearch: https://www.pyimagesearch.com/2015/12/28/increasing-raspberry-pi-fps-with-python-and-opencv/
class VideoStream:
    """Camera object that controls video streaming from the Picamera"""
    def __init__(self,resolution=(640,480),framerate=30):
        # Initialize the PiCamera and the camera image stream
        self.stream = cv2.VideoCapture(0)
        ret = self.stream.set(cv2.CAP_PROP_FOURCC, cv2.VideoWriter_fourcc(*'MJPG'))
        ret = self.stream.set(3,resolution[0])
        ret = self.stream.set(4,resolution[1])
            
        # Read first frame from the stream
        (self.grabbed, self.frame) = self.stream.read()

	# Variable to control when the camera is stopped
        self.stopped = False

    def start(self):
	# Start the thread that reads frames from the video stream
        Thread(target=self.update,args=()).start()
        return self

    def update(self):
        # Keep looping indefinitely until the thread is stopped
        while True:
            # If the camera is stopped, stop the thread
            if self.stopped:
                # Close camera resources
                self.stream.release()
                return

            # Otherwise, grab the next frame from the stream
            (self.grabbed, self.frame) = self.stream.read()

    def read(self):
	# Return the most recent frame
        return self.frame

    def stop(self):
	# Indicate that the camera and thread should be stopped
        self.stopped = True

# Define and parse input arguments
parser = argparse.ArgumentParser()
parser.add_argument('--modeldir', help='Folder the .tflite file is located in',
                    required=True)
parser.add_argument('--graph', help='Name of the .tflite file, if different than detect.tflite',
                    default='detect.tflite')
parser.add_argument('--labels', help='Name of the labelmap file, if different than labelmap.txt',
                    default='labelmap.txt')
parser.add_argument('--threshold', help='Minimum confidence threshold for displaying detected objects',
                    default=0.5)
parser.add_argument('--resolution', help='Desired webcam resolution in WxH. If the webcam does not support the resolution entered, errors may occur.',
                    default='1280x720')
parser.add_argument('--edgetpu', help='Use Coral Edge TPU Accelerator to speed up detection',
                    action='store_true')

args = parser.parse_args()

MODEL_NAME = args.modeldir
GRAPH_NAME = args.graph
LABELMAP_NAME = args.labels
min_conf_threshold = float(args.threshold)
resW, resH = args.resolution.split('x')
imW, imH = int(resW), int(resH)
use_TPU = args.edgetpu

# Import TensorFlow libraries
# If tflite_runtime is installed, import interpreter from tflite_runtime, else import from regular tensorflow
# If using Coral Edge TPU, import the load_delegate library
pkg = importlib.util.find_spec('tflite_runtime')
if pkg:
    from tflite_runtime.interpreter import Interpreter
    if use_TPU:
        from tflite_runtime.interpreter import load_delegate
else:
    from tensorflow.lite.python.interpreter import Interpreter
    if use_TPU:
        from tensorflow.lite.python.interpreter import load_delegate

# If using Edge TPU, assign filename for Edge TPU model
if use_TPU:
    # If user has specified the name of the .tflite file, use that name, otherwise use default 'edgetpu.tflite'
    if (GRAPH_NAME == 'detect.tflite'):
        GRAPH_NAME = 'edgetpu.tflite'       

# Get path to current working directory
CWD_PATH = os.getcwd()

# Path to .tflite file, which contains the model that is used for object detection
PATH_TO_CKPT = os.path.join(CWD_PATH,MODEL_NAME,GRAPH_NAME)

# Path to label map file
PATH_TO_LABELS = os.path.join(CWD_PATH,MODEL_NAME,LABELMAP_NAME)

# Load the label map
with open(PATH_TO_LABELS, 'r') as f:
    labels = [line.strip() for line in f.readlines()]

# Have to do a weird fix for label map if using the COCO "starter model" from
# https://www.tensorflow.org/lite/models/object_detection/overview
# First label is '???', which has to be removed.
if labels[0] == '???':
    del(labels[0])

# Load the Tensorflow Lite model.
# If using Edge TPU, use special load_delegate argument
if use_TPU:
    interpreter = Interpreter(model_path=PATH_TO_CKPT,
                              experimental_delegates=[load_delegate('libedgetpu.so.1.0')])
    print(PATH_TO_CKPT)
else:
    interpreter = Interpreter(model_path=PATH_TO_CKPT)

interpreter.allocate_tensors()

# Get model details
input_details = interpreter.get_input_details()
output_details = interpreter.get_output_details()
height = input_details[0]['shape'][1]
width = input_details[0]['shape'][2]

floating_model = (input_details[0]['dtype'] == np.float32)

input_mean = 127.5
input_std = 127.5

# Initialize frame rate calculation
frame_rate_calc = 1
freq = cv2.getTickFrequency()

# Initialize video stream
videostream = VideoStream(resolution=(imW,imH),framerate=30).start()
time.sleep(1)

#for frame1 in camera.capture_continuous(rawCapture, format="bgr",use_video_port=True):
while True:

    # Start timer (for calculating frame rate)
    t1 = cv2.getTickCount()

    # Grab frame from video stream
    frame1 = videostream.read()

    # Acquire frame and resize to expected shape [1xHxWx3]
    frame = frame1.copy()
    frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    frame_resized = cv2.resize(frame_rgb, (width, height))
    input_data = np.expand_dims(frame_resized, axis=0)

    # Normalize pixel values if using a floating model (i.e. if model is non-quantized)
    if floating_model:
        input_data = (np.float32(input_data) - input_mean) / input_std

    # Perform the actual detection by running the model with the image as input
    interpreter.set_tensor(input_details[0]['index'],input_data)
    interpreter.invoke()

    # Retrieve detection results
    boxes = interpreter.get_tensor(output_details[0]['index'])[0] # Bounding box coordinates of detected objects
    classes = interpreter.get_tensor(output_details[1]['index'])[0] # Class index of detected objects
    scores = interpreter.get_tensor(output_details[2]['index'])[0] # Confidence of detected objects
    #num = interpreter.get_tensor(output_details[3]['index'])[0]  # Total number of detected objects (inaccurate and not needed)

    # Loop over all detections and draw detection box if confidence is above minimum threshold
    for i in range(len(scores)):
        if ((scores[i] > min_conf_threshold) and (scores[i] <= 1.0)):

            # Get bounding box coordinates and draw box
            # Interpreter can return coordinates that are outside of image dimensions, need to force them to be within image using max() and min()
            ymin = int(max(1,(boxes[i][0] * imH)))
            xmin = int(max(1,(boxes[i][1] * imW)))
            ymax = int(min(imH,(boxes[i][2] * imH)))
            xmax = int(min(imW,(boxes[i][3] * imW)))
            
            cv2.rectangle(frame, (xmin,ymin), (xmax,ymax), (10, 255, 0), 2)

            # Draw label
            object_name = labels[int(classes[i])] # Look up object name from "labels" array using class index
            label = '%s: %d%%' % (object_name, int(scores[i]*100)) # Example: 'person: 72%'
            labelSize, baseLine = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.7, 2) # Get font size
            label_ymin = max(ymin, labelSize[1] + 10) # Make sure not to draw label too close to top of window
            cv2.rectangle(frame, (xmin, label_ymin-labelSize[1]-10), (xmin+labelSize[0], label_ymin+baseLine-10), (255, 255, 255), cv2.FILLED) # Draw white box to put label text in
            cv2.putText(frame, label, (xmin, label_ymin-7), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 0, 0), 2) # Draw label text

    # Draw framerate in corner of frame
    cv2.putText(frame,'FPS: {0:.2f}'.format(frame_rate_calc),(30,50),cv2.FONT_HERSHEY_SIMPLEX,1,(255,255,0),2,cv2.LINE_AA)

    # All the results have been drawn on the frame, so it's time to display it.
    cv2.imshow('Object detector', frame)

    # Calculate framerate
    t2 = cv2.getTickCount()
    time1 = (t2-t1)/freq
    frame_rate_calc= 1/time1

    # Press 'q' to quit
    if cv2.waitKey(1) == ord('q'):
        break

# Clean up
cv2.destroyAllWindows()
videostream.stop()

That’s the code of the Webcam detect script.

Hey Rorque,

I don’t see anything in the script that trips your GPIO output, have you modified this to use the interpreter to trigger another script once an object has been detected, I’ve had a read through it but I couldn’t find any reference to this in the code that you’ve copied through.

Within that script there isnt any output code that ive tried, i can enter it here and show you which ones ive tried;

This was from a thread i found: (python - Tensorflow Object detection API : Turn a GPIO pin high when a specific class is detected - Stack Overflow)

import RPi.GPIO as GPIO

GPIO.setwarnings(False)
GPIO.setmode(GPIO.BOARD)
GPIO.setup(11, GPIO.OUT, initial=GPIO.LOW)
GPIO.setup(13, GPIO.OUT, initial=GPIO.LOW)

if object_detected_class==1:
    GPIO.output(11, GPIO.HIGH)

else: #other class
    GPIO.output(13, GPIO.HIGH)

or

from gpiozero import LED

led = LED(11) #pin you plug led
filter=0.3 #filter of scores if you decrease it program finds more item
selected_class=4 #you want to find class
isledhigh=False
...
    makeledhigh=False
    (boxes, scores, classes, num) = sess.run(
            [detection_boxes, detection_scores, detection_classes, num_detections],
            feed_dict={image_tensor: frame_expanded})
    for i in range(int(num[0])):
        if classes[i]==selected_class and scores[i]>=filter:
            makeledhigh=True
    if makeledhigh and !isledhigh:
        led.on()
        isledhigh=True
    if isledhigh and !makeledhigh:
        led.off()
        isledhigh=False

Hi Rorque,

That should be functional as long as object_detected_class is returning a boolean value. So how are you calling this code? And what does your script object_detected_class look like? This appears to be an example that was put together on stack overflow for a similar question.

Hey Rorque,

This is more to confirm its working, but as an alternative, just try using this slightly modified code for your webcam detect script.

######## Webcam Object Detection Using Tensorflow-trained Classifier #########
#
# Author: Evan Juras
# Date: 10/27/19
# Description: 
# This program uses a TensorFlow Lite model to perform object detection on a live webcam
# feed. It draws boxes and scores around the objects of interest in each frame from the
# webcam. To improve FPS, the webcam object runs in a separate thread from the main program.
# This script will work with either a Picamera or regular USB webcam.
#
# This code is based off the TensorFlow Lite image classification example at:
# https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/examples/python/label_image.py
#
# I added my own method of drawing boxes and labels using OpenCV.

# Import packages
import os
import argparse
import cv2
import numpy as np
import sys
import time
from threading import Thread
import importlib.util

#------------------------------------------------------------ADDED-------------------------------------------------------------
import RPi.GPIO as GPIO

GPIO.setwarnings(False)
GPIO.setmode(GPIO.BOARD)
GPIO.setup(11, GPIO.OUT, initial=GPIO.LOW)

    

# Define VideoStream class to handle streaming of video from webcam in separate processing thread
# Source - Adrian Rosebrock, PyImageSearch: https://www.pyimagesearch.com/2015/12/28/increasing-raspberry-pi-fps-with-python-and-opencv/
class VideoStream:
    """Camera object that controls video streaming from the Picamera"""
    def __init__(self,resolution=(640,480),framerate=30):
        # Initialize the PiCamera and the camera image stream
        self.stream = cv2.VideoCapture(0)
        ret = self.stream.set(cv2.CAP_PROP_FOURCC, cv2.VideoWriter_fourcc(*'MJPG'))
        ret = self.stream.set(3,resolution[0])
        ret = self.stream.set(4,resolution[1])
            
        # Read first frame from the stream
        (self.grabbed, self.frame) = self.stream.read()

	# Variable to control when the camera is stopped
        self.stopped = False

    def start(self):
	# Start the thread that reads frames from the video stream
        Thread(target=self.update,args=()).start()
        return self

    def update(self):
        # Keep looping indefinitely until the thread is stopped
        while True:
            # If the camera is stopped, stop the thread
            if self.stopped:
                # Close camera resources
                self.stream.release()
                return

            # Otherwise, grab the next frame from the stream
            (self.grabbed, self.frame) = self.stream.read()

    def read(self):
	# Return the most recent frame
        return self.frame

    def stop(self):
	# Indicate that the camera and thread should be stopped
        self.stopped = True

# Define and parse input arguments
parser = argparse.ArgumentParser()
parser.add_argument('--modeldir', help='Folder the .tflite file is located in',
                    required=True)
parser.add_argument('--graph', help='Name of the .tflite file, if different than detect.tflite',
                    default='detect.tflite')
parser.add_argument('--labels', help='Name of the labelmap file, if different than labelmap.txt',
                    default='labelmap.txt')
parser.add_argument('--threshold', help='Minimum confidence threshold for displaying detected objects',
                    default=0.5)
parser.add_argument('--resolution', help='Desired webcam resolution in WxH. If the webcam does not support the resolution entered, errors may occur.',
                    default='1280x720')
parser.add_argument('--edgetpu', help='Use Coral Edge TPU Accelerator to speed up detection',
                    action='store_true')

args = parser.parse_args()

MODEL_NAME = args.modeldir
GRAPH_NAME = args.graph
LABELMAP_NAME = args.labels
min_conf_threshold = float(args.threshold)
resW, resH = args.resolution.split('x')
imW, imH = int(resW), int(resH)
use_TPU = args.edgetpu

# Import TensorFlow libraries
# If tflite_runtime is installed, import interpreter from tflite_runtime, else import from regular tensorflow
# If using Coral Edge TPU, import the load_delegate library
pkg = importlib.util.find_spec('tflite_runtime')
if pkg:
    from tflite_runtime.interpreter import Interpreter
    if use_TPU:
        from tflite_runtime.interpreter import load_delegate
else:
    from tensorflow.lite.python.interpreter import Interpreter
    if use_TPU:
        from tensorflow.lite.python.interpreter import load_delegate

# If using Edge TPU, assign filename for Edge TPU model
if use_TPU:
    # If user has specified the name of the .tflite file, use that name, otherwise use default 'edgetpu.tflite'
    if (GRAPH_NAME == 'detect.tflite'):
        GRAPH_NAME = 'edgetpu.tflite'       

# Get path to current working directory
CWD_PATH = os.getcwd()

# Path to .tflite file, which contains the model that is used for object detection
PATH_TO_CKPT = os.path.join(CWD_PATH,MODEL_NAME,GRAPH_NAME)

# Path to label map file
PATH_TO_LABELS = os.path.join(CWD_PATH,MODEL_NAME,LABELMAP_NAME)

# Load the label map
with open(PATH_TO_LABELS, 'r') as f:
    labels = [line.strip() for line in f.readlines()]

# Have to do a weird fix for label map if using the COCO "starter model" from
# https://www.tensorflow.org/lite/models/object_detection/overview
# First label is '???', which has to be removed.
if labels[0] == '???':
    del(labels[0])

# Load the Tensorflow Lite model.
# If using Edge TPU, use special load_delegate argument
if use_TPU:
    interpreter = Interpreter(model_path=PATH_TO_CKPT,
                              experimental_delegates=[load_delegate('libedgetpu.so.1.0')])
    print(PATH_TO_CKPT)
else:
    interpreter = Interpreter(model_path=PATH_TO_CKPT)

interpreter.allocate_tensors()

# Get model details
input_details = interpreter.get_input_details()
output_details = interpreter.get_output_details()
height = input_details[0]['shape'][1]
width = input_details[0]['shape'][2]

floating_model = (input_details[0]['dtype'] == np.float32)

input_mean = 127.5
input_std = 127.5

# Initialize frame rate calculation
frame_rate_calc = 1
freq = cv2.getTickFrequency()

# Initialize video stream
videostream = VideoStream(resolution=(imW,imH),framerate=30).start()
time.sleep(1)

#for frame1 in camera.capture_continuous(rawCapture, format="bgr",use_video_port=True):
while True:

    # Start timer (for calculating frame rate)
    t1 = cv2.getTickCount()

    # Grab frame from video stream
    frame1 = videostream.read()

    # Acquire frame and resize to expected shape [1xHxWx3]
    frame = frame1.copy()
    frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    frame_resized = cv2.resize(frame_rgb, (width, height))
    input_data = np.expand_dims(frame_resized, axis=0)

    # Normalize pixel values if using a floating model (i.e. if model is non-quantized)
    if floating_model:
        input_data = (np.float32(input_data) - input_mean) / input_std

    # Perform the actual detection by running the model with the image as input
    interpreter.set_tensor(input_details[0]['index'],input_data)
    interpreter.invoke()

    # Retrieve detection results
    boxes = interpreter.get_tensor(output_details[0]['index'])[0] # Bounding box coordinates of detected objects
    classes = interpreter.get_tensor(output_details[1]['index'])[0] # Class index of detected objects
    scores = interpreter.get_tensor(output_details[2]['index'])[0] # Confidence of detected objects
    #num = interpreter.get_tensor(output_details[3]['index'])[0]  # Total number of detected objects (inaccurate and not needed)

    # Loop over all detections and draw detection box if confidence is above minimum threshold
    for i in range(len(scores)):
        if ((scores[i] > min_conf_threshold) and (scores[i] <= 1.0)):

            # Get bounding box coordinates and draw box
            # Interpreter can return coordinates that are outside of image dimensions, need to force them to be within image using max() and min()
            ymin = int(max(1,(boxes[i][0] * imH)))
            xmin = int(max(1,(boxes[i][1] * imW)))
            ymax = int(min(imH,(boxes[i][2] * imH)))
            xmax = int(min(imW,(boxes[i][3] * imW)))
            
            cv2.rectangle(frame, (xmin,ymin), (xmax,ymax), (10, 255, 0), 2)

            # Draw label
            object_name = labels[int(classes[i])] # Look up object name from "labels" array using class index
            label = '%s: %d%%' % (object_name, int(scores[i]*100)) # Example: 'person: 72%'
            labelSize, baseLine = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.7, 2) # Get font size
            label_ymin = max(ymin, labelSize[1] + 10) # Make sure not to draw label too close to top of window
            cv2.rectangle(frame, (xmin, label_ymin-labelSize[1]-10), (xmin+labelSize[0], label_ymin+baseLine-10), (255, 255, 255), cv2.FILLED) # Draw white box to put label text in
            cv2.putText(frame, label, (xmin, label_ymin-7), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 0, 0), 2) # Draw label text

    # Draw framerate in corner of frame
    cv2.putText(frame,'FPS: {0:.2f}'.format(frame_rate_calc),(30,50),cv2.FONT_HERSHEY_SIMPLEX,1,(255,255,0),2,cv2.LINE_AA)

    # All the results have been drawn on the frame, so it's time to display it.
    cv2.imshow('Object detector', frame)

    #------------------------------------------------------------ADDED-------------------------------------------------------------
    #you will need to edit this object name check to something that actually exists in the TensorFlow model
    if object_name == "person":
        GPIO.output(11, GPIO.HIGH)
    else:
        GPIO.output(11, GPIO.LOW)



    # Calculate framerate
    t2 = cv2.getTickCount()
    time1 = (t2-t1)/freq
    frame_rate_calc= 1/time1

    # Press 'q' to quit
    if cv2.waitKey(1) == ord('q'):
        break

# Clean up
cv2.destroyAllWindows()
videostream.stop()

I have put --ADDED-- above the added sections of the script. When it detects a “person” it will trigger GPIO_11 High, if it’s not detecting anything it will trigger low. Just remember it does this once per frame of detection so even at the few FPS you get from the Pi it can still trigger fairly often.

Hope this helps!

Cheers for that!

I tried to run it and got the following error;

Traceback (most recent call last):
File “TFLite_detection_webcam.py”, line 27, in
import RPi.GPIO as GPIO
ModuleNotFoundError: No module named ‘RPi’

edit
I have installed RPi.GPIO, restarting the PI and going to see if that does anything.

edit 2
Restarted and still no change.
ModuleNotFoundError: No module named ‘RPi’

Hey Rorque,

You can likely resolve that by just putting the RPi GPIO trigger script in another file and running it from the webcam detect script instead (like you have done with that second lot of scripts you have sent through to Bryce).

Hey!

the scripts above were just the examples that I couldnt get working. I might just have to see if i can figure out the RPi.GPIO error and hopefully everything goes well from there.