app.py

import gradio as gr
from dotenv import load_dotenv
from roboflow import Roboflow
import tempfile
import os
import cv2
import numpy as np
from dds_cloudapi_sdk import Config, Client
from dds_cloudapi_sdk.tasks.dinox import DinoxTask
from dds_cloudapi_sdk.tasks.types import DetectionTarget
from dds_cloudapi_sdk import TextPrompt
import supervision as sv

# ========== Configuration ==========
load_dotenv()

# Roboflow Config
rf_api_key = os.getenv("ROBOFLOW_API_KEY")
workspace = os.getenv("ROBOFLOW_WORKSPACE")
project_name = os.getenv("ROBOFLOW_PROJECT")
model_version = int(os.getenv("ROBOFLOW_MODEL_VERSION"))

# DINO-X Config
DINOX_API_KEY = os.getenv("DINO_X_API_KEY")
DINOX_PROMPT = "beverage . bottle"  # Customize for competitor products

# Initialize Models
rf = Roboflow(api_key=rf_api_key)
project = rf.workspace(workspace).project(project_name)
yolo_model = project.version(model_version).model
dinox_config = Config(DINOX_API_KEY)
dinox_client = Client(dinox_config)

# ========== Combined Detection Function ==========
def detect_combined(image):
    with tempfile.NamedTemporaryFile(delete=False, suffix=".jpg") as temp_file:
        image.save(temp_file, format="JPEG")
        temp_path = temp_file.name
    try:
        # [1] YOLO: Detect Nestlé Products
        yolo_pred = yolo_model.predict(temp_path, confidence=60, overlap=80).json()
        nestle_class_count = {}
        nestle_boxes = []
        for pred in yolo_pred['predictions']:
            class_name = pred['class']
            nestle_class_count[class_name] = nestle_class_count.get(class_name, 0) + 1
            nestle_boxes.append((pred['x'], pred['y'], pred['width'], pred['height']))
        total_nestle = sum(nestle_class_count.values())

        # [2] DINO-X: Detect Competitor Products
        image_url = dinox_client.upload_file(temp_path)
        task = DinoxTask(
            image_url=image_url,
            prompts=[TextPrompt(text=DINOX_PROMPT)],
            bbox_threshold=0.25,
            targets=[DetectionTarget.BBox]
        )
        dinox_client.run_task(task)
        dinox_pred = task.result.objects

        # Filter & Count Competitors
        competitor_class_count = {}
        competitor_boxes = []
        for obj in dinox_pred:
            dinox_box = obj.bbox
            if not is_overlap(dinox_box, nestle_boxes):
                class_name = obj.category.strip().lower()
                competitor_class_count[class_name] = competitor_class_count.get(class_name, 0) + 1
                competitor_boxes.append({
                    "class": class_name,
                    "box": dinox_box,
                    "confidence": obj.score
                })
        total_competitor = sum(competitor_class_count.values())

        # [3] Format Output
        result_text = "Product Nestle\n\n"
        for class_name, count in nestle_class_count.items():
            result_text += f"{class_name}: {count}\n"
        result_text += f"\nTotal Product Nestle: {total_nestle}\n\n"
        result_text += "Competitor Products\n\n"
        if competitor_class_count:
            for class_name, count in competitor_class_count.items():
                result_text += f"{class_name}: {count}\n"
        else:
            result_text += "No competitors detected\n"
        result_text += f"\nTotal Competitor: {total_competitor}"

        # [4] Visualization
        img = cv2.imread(temp_path)
        # Nestlé (Green)
        for pred in yolo_pred['predictions']:
            x, y, w, h = pred['x'], pred['y'], pred['width'], pred['height']
            cv2.rectangle(img, (int(x-w/2), int(y-h/2)), (int(x+w/2), int(y+h/2)), (0, 255, 0), 2)
            cv2.putText(img, pred['class'], (int(x-w/2), int(y-h/2-10)), 
                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
        # Competitors (Red)
        for comp in competitor_boxes:
            x1, y1, x2, y2 = comp['box']
            cv2.rectangle(img, (int(x1), int(y1)), (int(x2), int(y2)), (0, 0, 255), 2)
            cv2.putText(img, f"{comp['class']} {comp['confidence']:.2f}", 
                        (int(x1), int(y1-10)), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 2)
        output_path = "/tmp/combined_output.jpg"
        cv2.imwrite(output_path, img)
        return output_path, result_text
    except Exception as e:
        return temp_path, f"Error: {str(e)}"
    finally:
        if os.path.exists(temp_path):
            os.remove(temp_path)

# ========== Overlap Detection Function ==========
def is_overlap(box1, boxes2, threshold=0.3):
    x1_min, y1_min, x1_max, y1_max = box1
    for b2 in boxes2:
        x2, y2, w2, h2 = b2
        x2_min = x2 - w2/2
        x2_max = x2 + w2/2
        y2_min = y2 - h2/2
        y2_max = y2 + h2/2
        dx = min(x1_max, x2_max) - max(x1_min, x2_min)
        dy = min(y1_max, y2_max) - max(y1_min, y2_min)
        if (dx >= 0) and (dy >= 0):
            area_overlap = dx * dy
            area_box1 = (x1_max - x1_min) * (y1_max - y1_min)
            if area_overlap / area_box1 > threshold:
                return True
    return False

# ========== Video Detection Function ==========
def detect_objects_in_video(video_path):
    temp_output_path = "/tmp/output_video.mp4"
    temp_frames_dir = tempfile.mkdtemp()
    try:
        video = cv2.VideoCapture(video_path)
        frame_rate = int(video.get(cv2.CAP_PROP_FPS))
        frame_width = int(video.get(cv2.CAP_PROP_FRAME_WIDTH))
        frame_height = int(video.get(cv2.CAP_PROP_FRAME_HEIGHT))
        frame_size = (frame_width, frame_height)
        fourcc = cv2.VideoWriter_fourcc(*'mp4v')
        output_video = cv2.VideoWriter(temp_output_path, fourcc, frame_rate, frame_size)
        frame_index = 0
        while True:
            ret, frame = video.read()
            if not ret:
                break
            frame_path = os.path.join(temp_frames_dir, f"frame_{frame_index}.jpg")
            cv2.imwrite(frame_path, frame)
            predictions = yolo_model.predict(frame_path, confidence=60, overlap=80).json()
            class_count = {}
            for prediction in predictions['predictions']:
                class_name = prediction['class']
                class_count[class_name] = class_count.get(class_name, 0) + 1
            text_offset = 30
            y_position = 30
            for class_name, count in class_count.items():
                cv2.putText(frame, f"{class_name}: {count}", (10, y_position), 
                            cv2.FONT_HERSHEY_SIMPLEX, 0.8, (255, 255, 255), 2, cv2.LINE_AA)
                y_position += text_offset
            for prediction in predictions['predictions']:
                x, y, w, h = prediction['x'], prediction['y'], prediction['width'], prediction['height']
                cv2.rectangle(frame, (int(x - w/2), int(y - h/2)), (int(x + w/2), int(y + h/2)), (0, 255, 0), 2)
                cv2.putText(frame, prediction['class'], (int(x - w/2), int(y - h/2 - 10)), 
                            cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
            output_video.write(frame)
            frame_index += 1
        video.release()
        output_video.release()
        return temp_output_path
    except Exception as e:
        return None, f"An error occurred: {e}"

# ========== Gradio Interface ==========
with gr.Blocks(theme=gr.themes.Base(primary_hue="teal", secondary_hue="teal", neutral_hue="slate")) as iface:
    gr.Markdown("""
    <div style="text-align: center;">
        <h1>NESTLE - STOCK COUNTING</h1>
    </div>
    """)
    with gr.Row():
        with gr.Column():
            input_image = gr.Image(type="pil", label="Input Image")
            detect_image_button = gr.Button("Detect Image")
            output_image = gr.Image(label="Detect Object")
            output_text = gr.Textbox(label="Counting Object")
            detect_image_button.click(
                fn=detect_combined,
                inputs=input_image,
                outputs=[output_image, output_text]
            )
        with gr.Column():
            input_video = gr.Video(label="Input Video")
            detect_video_button = gr.Button("Detect Video")
            output_video = gr.Video(label="Output Video")
            detect_video_button.click(
                fn=detect_objects_in_video,
                inputs=input_video,
                outputs=[output_video]
            )
iface.launch()