Easy Raspberry pi 5 and OpenCV based object detection and sorting using python

Hello tech enthusiasts

Ever wondered how speed cameras caught over speeding and recognized the owner and send ticket for it, how object tracking robots and autonomous vehicles work?                                                                  

They all have one common technology in them image processing and computer vision using which we have made object detection and sorting. 

And in this blog, we will demonstrate how to setup hardware, flash raspberry pi 5, setup open cv , write code for it. So, whether you're a beginner or an experienced programmer, this guide will walk you through the process step-by-step. Let's dive in! 

 What is Open CV?

It is an open-source software library for computer vision and machine learning is called OpenCV (Open-Source Computer Vision Library). It has over 2500 efficient algorithms for a variety of computer vision tasks, such as image processing, face recognition, and object identification.  It provides various useful tools and functions and supports various programing language

Key features

Image processing: - open cv allows to perform various task such as edge detection, filtering and many more

Computer vision: - it includes object detection, feature detection, and image stitching

Object tracking: - it provides object tracing which enables video surveillance and augmented reality.

Gesture Recognition: -  it interprets human gestures using sensors and algorithms to enable natural interaction with devices.

Machine Learning: -

If you want to learn more you can view official website of open CV-  https://docs.opencv.org

Why Raspberry Pi 5? 

The Raspberry Pi 5 is a small, affordable on board computer that can be  use for various programming and electronics projects. The Raspberry Pi 5, with its powerful processing capabilities, is perfect for running computer vision applications like this one we are building with OpenCV. 

Read more here - https://www.raspberrypi.com/products/raspberry-pi-5/

Hardware Requirements

• Raspberry pi 5
• Raspberry pi 5 case /active cooler  
• Power adapter for raspberry pi 5
• Micro-HDMI (Male) to Standard HDMI (Male)
• Monitor /screen
• Keyboard
• Mouse
• Camera
• Micro SD card
• SD Card adapter/Card reader
• LED
• Servo motor

Software requirement

• Thonny
• visual studio code

Setup

• Hardware setup for Raspberry Pi 5
• Visual studio code installation
• OpenCV installation

Setting Up Raspberry Pi 5 

Before we start coding, let's set up Raspberry Pi 5: 

You can read this for detailed process - https://robu.in/set-up-raspberry-pi-5-step-by-step-beginners-guide/ 

Step 1: Install the Operating System 

1. Download Raspberry Pi OS: Visit the official Raspberry Pi website and download the latest version of Raspberry Pi OS. 

2. Flash the OS: Use a tool like Balena Etcher to flash the OS image onto a microSD card. 

3. Boot the Raspberry Pi: Insert the microSD card into your Raspberry Pi, connect it to a monitor, keyboard, and mouse, and power it up. 

OpenCV installation

• Update and Upgrade: 

sudo apt-get update && sudo apt-get upgrade -y

Python library -

$ sudo apt-get install python3-opencv

 If you need detail article you can refer – https://qengineering.eu/install%20opencv%20on%20raspberry%20pi%205.html

To get Pre trained model and libraries refer this-https://core-electronics.com.au/guides/object-identify-raspberry-pi/

Code

Get complete code file here-https://github.com/Robu-In/Object-Recognition

 import cv2 
 import asyncio 
 import gpiod 
 import time 

  

LED_PIN = 17 
 chip = gpiod.Chip('gpiochip4') 
 led_line = chip.get_line(LED_PIN) 
 led_line.request(consumer="LED", type=gpiod.LINE_REQ_DIR_OUT) 

  

#thres = 0.45 # Threshold to detect object 
 classNames = [] 
 classFile = "Files/coco.names" 
 with open(classFile,"rt") as f: 
     classNames = f.read().rstrip("\n").split("\n") 

  

configPath = "Files/ssd_mobilenet_v3_large_coco_2020_01_14.pbtxt" 
 weightsPath = "Files/frozen_inference_graph.pb" 

  

net = cv2.dnn_DetectionModel(weightsPath,configPath) 
 net.setInputSize(320,320) 
 net.setInputScale(1.0/ 127.5) 
 net.setInputMean((127.5, 127.5, 127.5)) 
 net.setInputSwapRB(True) 

  

def getObjects(img, thres, nms, draw=True, objects=[]): 
     classIds, confs, bbox = net.detect(img,confThreshold=thres,nmsThreshold=nms) 
     if len(objects) == 0: objects = classNames 
     objectInfo =[] 
     if len(classIds) != 0: 
         for classId, confidence,box in zip(classIds.flatten(),confs.flatten(),bbox): 
             className = classNames[classId - 1] 
             if className in objects: 
                 objectInfo.append([box,className]) 
                 if (draw): 
                     cv2.rectangle(img,box,color=(0,255,0),thickness=2) 
                     cv2.putText(img,classNames[classId-1].upper(),(box[0]+10,box[1]+30), 
                     cv2.FONT_HERSHEY_COMPLEX,1,(0,255,0),2) 

  

    return img,objectInfo 

  

def set_led(value): 
     global led_line 
     led_line.set_value(value) 

  

async def main(): 
     cap = cv2.VideoCapture(0) 

    cap.set(3,640) 
     cap.set(4,480) 

  

    while True: 
         success, img = cap.read() 
         result, objectInfo = getObjects(img,0.45,0.2, objects=['scissors']) 
         # print(objectInfo) 

  

        found = any('scissors' in sublist for sublist in objectInfo) 
         if found: 
             set_led(1) 
             time.sleep(2) 
             set_led(0) 
             time.sleep(1) 

  

        cv2.imshow("Output",img) 
         cv2.waitKey(1) 

  

if __name__ == "__main__": 
     asyncio.run(main())

Results

Here we have updated code to detect scissors and given out the detection signal to GPIO pin

And as scissor is detected Red LED is turned ON.

Similarly we can update code to detect any object and further program Pi to sort it accordingly.

Conclusion

We have successfully set up Raspberry Pi 5 with OpenCV for object detection and sorting. This project showcases the powerful capabilities of combining computer vision with physical computing. From here, we can expand the project by adding more complex sorting mechanisms, integrating additional sensors, or applying advanced machine learning techniques. 

Watch related videos- https://youtu.be/jqFQkm45vJI

Related blog- https://robu.in/set-up-raspberry-pi-5-step-by-step-beginners-guide/