count_boxes.py

import gradio as gr
import cv2
import requests
import os

from ultralytics import YOLO

# Define the directory containing the YOLO model files (*.pt)
model_weights_dir = 'best_weights'

# List all "*.pt" files in the specified directory
model_paths = [os.path.join(model_weights_dir, filename) for filename in os.listdir(model_weights_dir) if filename.endswith('.pt')]

# Initialize YOLO models based on the discovered model paths
models = [YOLO(model_path) for model_path in model_paths]

# Extract model names from paths (remove directory and ".pt" extension)
model_names = [os.path.splitext(os.path.basename(model_path))[0] for model_path in model_paths]

path = [['plot.JPG']]

def show_preds_image(image_path, selection):
    image = cv2.imread(image_path)
    outputs = models[selection].predict(source=image_path)
    results = outputs[0].cpu().numpy()
    for i, det in enumerate(results.boxes.xyxy):
        cv2.rectangle(
            image,
            (int(det[0]), int(det[1])),
            (int(det[2]), int(det[3])),
            color=(0, 0, 255),
            thickness=2,
            lineType=cv2.LINE_AA
        )

    box_count = len(results)   #counting the number of boxes detected
    return cv2.cvtColor(image, cv2.COLOR_BGR2RGB), box_count


inputs = [
    gr.components.Image(type="filepath", label="Input Image"),
    gr.components.Dropdown(choices=list(zip(model_names, range(len(models)))), label="Select Model"),
]

outputs_image = [
    gr.components.Image(type="numpy", label="Output Image"),
    gr.components.Textbox(label="Box Count", type="auto"),
]


# Create the Gradio interface
interface_image = gr.Interface(
    fn=show_preds_image,
    inputs=inputs,
    outputs=outputs_image,
    title="Paddy Growth Stage Recognition",
    description="Select an image and a YOLO model to detect the growth stage",
    examples=path,
    cache_examples=False,
)


gr.TabbedInterface(
    [interface_image],
    tab_names=['Image inference']
).queue().launch()