Robotflow.py

# pip install inference-cli --user
# pip install supervision
# pip install openv-python
# sudo apt-get update
# sudo apt-get install -y libgl1
# pip install pillow


from inference_sdk import InferenceHTTPClient
import cv2
import supervision as sv
from tkinter import filedialog
from tkinter import Tk
from PIL import Image

# Initiate default value
image = cv2.imread("/workspaces/Green-AI-Event/resources/car.jpg")
model_id = "vehicle-detection-bz0yu/4"


# Configure client
client = InferenceHTTPClient(
    api_url="https://detect.roboflow.com",  # route for local inference server
    api_key="api key",  # api key for your workspace
)

# Run inference
result = client.infer(image, model_id=model_id)

# Load results into Supervision Detection API
detections = sv.Detections.from_inference(result)

# Create Supervision annotators
box_annotator = sv.BoxAnnotator()
label_annotator = sv.LabelAnnotator()

# Extract labels array from inference results
labels = [p['class'] for p in result['predictions']]

# Apply results to image using Supervision annotators
annotated_image = box_annotator.annotate(scene=image, detections=detections)
annotated_image = label_annotator.annotate(
    scene=annotated_image, detections=detections, labels=labels
)

# Write annotated image to file or display image
with sv.ImageSink(target_dir_path="./results/", overwrite=True) as sink:
    sink.save_image(annotated_image)
# or sv.plot_image(annotated_image)

# Display the annotated image in a pop-up window using OpenCV
image_path = ("/workspaces/Green-AI-Event/results/image_00000.png")
# Change this to your specific path

# Open the image in the default viewer
img = Image.open(image_path)
img.show()




# Local Port Version

# model_id = "vehicle-detection-bz0yu/4"


# def select_image():
#     # Hide root window
#     root = Tk()
#     root.withdraw()

#     # Open file dialog to select image
#     image_file = filedialog.askopenfilenames(
#         title="Select File",
#         filetypes=[("Image Files", "*.jpg;*.jpeg;*.png;*.bmp;*.tiff")]
#     )
#     return image_file


# # Select an image from the file system
# image_file = select_image()
# if not image_file:
#     print("No image selected.")
# else:
#     # Get the first selected image file from the tuple
#     image_path = image_file[0]  # Use [0] to get the first file
#     print(image_path)
#     image = cv2.imread(image_path)

# # Configure client
# client = InferenceHTTPClient(
#     api_url="https://detect.roboflow.com",  # route for local inference server
#     api_key="<API-KEY>",  # api key for your workspace
# )

# # Run inference
# result = client.infer(image, model_id=model_id)

# # Load results into Supervision Detection API
# detections = sv.Detections.from_inference(result)

# # Create Supervision annotators
# box_annotator = sv.BoxAnnotator()
# label_annotator = sv.LabelAnnotator()

# # Extract labels array from inference results
# labels = [p['class'] for p in result['predictions']]

# # Apply results to image using Supervision annotators
# annotated_image = box_annotator.annotate(scene=image, detections=detections)
# annotated_image = label_annotator.annotate(
#     scene=annotated_image, detections=detections, labels=labels
# )

# # Write annotated image to file or display image
# with sv.ImageSink(target_dir_path="./results/", overwrite=True) as sink:
#     sink.save_image(annotated_image)
# # or sv.plot_image(annotated_image)

# # Display the annotated image in a pop-up window using OpenCV
# image_path = ("<Your Working Directory>" + "/results/image_00000.png")
# # Change this to your specific path

# # Open the image in the default viewer
# img = Image.open(image_path)
# img.show()