How to create a Python script for real-time vehicle detection and counting?

To create a Python script for real-time vehicle detection and counting, we are going to use the OpenCV library which is a high-performance image processing library. We will also be using the yolov3 model for the detection stage. It's a model trained on the COCO dataset and can detect up to 80 classes of objects.

Installation
Before we begin, ensure that you have OpenCV, Numpy, Imutils, and Dlib installed. You can install these packages using pip.

pip install opencv-python-headless numpy imutils dlib

Also, download the YOLO weight and configuration file.

• yolov3.cfg (https://github.com/pjreddie/darknet/blob/master/cfg/yolov3.cfg)
• yolov3.weights (https://pjreddie.com/media/files/yolov3.weights)

Here is a step-by-step guide.

Step 1: Import Required Libraries

import cv2 
import numpy as np
import imutils

Step 2: Load YOLO
We will load the yolov3 model.

net = cv2.dnn.readNet("yolov3.weights", "yolov3.cfg")
layer_names = net.getLayerNames()
output_layers = [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]

Step 3: Capture Video from Webcam

cap = cv2.VideoCapture(0)
font = cv2.FONT_HERSHEY_PLAIN
starting_time = time.time()
frame_id = 0

Step 4: Inside a Loop, Read Images from the Webcam

while True:
    _, frame = cap.read() 
    frame_id += 1
    height, width, channels = frame.shape
    # Detecting objects
    blob = cv2.dnn.blobFromImage(frame, 0.00392, (320, 320), (0, 0, 0), True, crop=False)
    net.setInput(blob)
    outs = net.forward(output_layers)

Step 5: Show Information on the Screen

class_ids = []
confidences = []
boxes = []
for out in outs:
    for detection in out:
        scores = detection[5:]
        class_id = np.argmax(scores)
        confidence = scores[class_id]
        if confidence > 0.2:
            # Object detected
            center_x = int(detection[0] * width)
            center_y = int(detection[1] * height)
            w = int(detection[2] * width)
            h = int(detection[3] * height)
            # Rectangle coordinates
            x = int(center_x - w / 2)
            y = int(center_y - h / 2)
            boxes.append([x, y, w, h])
            confidences.append(float(confidence))
            class_ids.append(class_id)

Step 6: Perform the Object Detection Inside the While Loop

indexes = cv2.dnn.NMSBoxes(boxes, confidences, 0.8, 0.3)
# The NMSBoxes function removes all the predicted bounding boxes that have a detection probability less than the threshold.

Step 7: Show the Output in a Window

for i in range(len(boxes)):
    if i in indexes:
        x, y, w, h = boxes[i]
        label = str(classes[class_ids[i]])
        confidence = confidences[i]
        color = colors[class_ids[i]]
        cv2.rectangle(frame, (x, y), (x + w, y + h), color, 2)
        cv2.putText(frame, label + " " + str(round(confidence, 2)), (x, y + 30), font, 1, color, 3)
    cv2.imshow("Image", frame)

Step 8: Break the Loop

key = cv2.waitKey(1) 
if key & 0xFF == ord('q'): 
    break
cap.release() 
cv2.destroyAllWindows()

After running this script, a window should open accessing your webcam where the script will detect the vehicles in your area in real time. It will count the number of vehicles as well with vehicle classes taken from the coco datasets.

This is a simplified vehicle detection script. You can further enhance it by filtering by cars and trucks only and also storing the number of vehicles detected. For real world scenarios, you need a more advanced way of counting vehicles over a period.