app.py

import os
from pathlib import Path
from typing import List, Union
from PIL import Image
import ezdxf.units
import numpy as np
import torch
from torchvision import transforms
from ultralytics import YOLOWorld, YOLO
from ultralytics.engine.results import Results
from ultralytics.utils.plotting import save_one_box
from transformers import AutoModelForImageSegmentation
import cv2
import ezdxf
import gradio as gr
import gc
from scalingtestupdated import calculate_scaling_factor
from scipy.interpolate import splprep, splev
from scipy.ndimage import gaussian_filter1d

# --- DOCUMENTATION STRINGS (Drawer Detection App) ---

GUIDELINE_SETUP = """
## 1. Quick Start Guide: Setup and DXF Generation

This application analyzes an image of items inside a drawer, calculates scaling, and outputs a manufacturing-ready DXF file with offsets applied.

1.  **Upload Image:** Upload a clear image of the drawer area, ensuring the items and the scaling reference box are visible.
2.  **Set Offset:** Enter the desired offset value in **inches**. This determines the clearance around the contour (e.g., 0.075 inches is the default).
3.  **Run:** Click the **"Submit"** button (or run using an example).
4.  **Review & Download:** Review the resulting images (Contoured Output, Outlines, Mask) and download the generated **DXF file**.
"""

GUIDELINE_INPUT = """
## 2. Expected Inputs and Preprocessing

| Input Field | Purpose | Requirement |
| :--- | :--- | :--- |
| **Input Image** | A high-resolution image of the drawer containing the objects to be contoured. | Must show the items and the reference scaling box clearly. |
| **Offset value (inches)** | The physical distance (clearance) to be added around the detected contours for manufacturing tolerance. | Input must be a positive number (float). Default is 0.075 inches. |


"""

GUIDELINE_OUTPUT = """
## 3. Expected Outputs (Manufacturing Results)

The application provides five key outputs:

1.  **Ouput Image:** The original cropped drawer image overlaid with the final, offset contours (blue lines).
2.  **Outlines of Objects:** A grayscale image showing only the final, smoothed contour lines.
3.  **DXF file (Downloadable):** The primary output. This file contains scaled 2D spline geometry (in inches) based on the calculated contours, ready for CAD or CNC machines.
4.  **Mask:** The raw, dilated binary mask used to generate the contours.
5.  **Scaling Factor (Textbox):** The calculated ratio (in pixels per inch) used to accurately convert pixel dimensions into real-world units for the DXF file.
"""
# ----------------------------------------------------
# END GUIDELINE DEFINITIONS
# ----------------------------------------------------


birefnet = AutoModelForImageSegmentation.from_pretrained(
    "zhengpeng7/BiRefNet", trust_remote_code=True
)

device = "cpu"
torch.set_float32_matmul_precision(["high", "highest"][0])

birefnet.to(device)
birefnet.eval()
transform_image = transforms.Compose(
    [
        transforms.Resize((1024, 1024)),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),
    ]
)


def yolo_detect(
    image: Union[str, Path, int, Image.Image, list, tuple, np.ndarray, torch.Tensor],
    classes: List[str],
) -> np.ndarray:
    drawer_detector = YOLOWorld("yolov8x-worldv2.pt")
    drawer_detector.set_classes(classes)
    results: List[Results] = drawer_detector.predict(image)
    boxes = []
    for result in results:
        boxes.append(
            save_one_box(result.cpu().boxes.xyxy, im=result.orig_img, save=False)
        )

    del drawer_detector
    gc.collect() # Ensure memory is cleared

    return boxes[0]


def remove_bg(image: np.ndarray) -> np.ndarray:
    image = Image.fromarray(image)
    input_images = transform_image(image).unsqueeze(0).to("cpu")

    # Prediction
    with torch.no_grad():
        preds = birefnet(input_images)[-1].sigmoid().cpu()
    pred = preds[0].squeeze()

    # Show Results
    pred_pil: Image = transforms.ToPILImage()(pred)
    # Scale proportionally with max length to 1024 for faster showing
    scale_ratio = 1024 / max(image.size)
    scaled_size = (int(image.size[0] * scale_ratio), int(image.size[1] * scale_ratio))

    return np.array(pred_pil.resize(scaled_size))


def make_square(img: np.ndarray):
    # Get dimensions
    height, width = img.shape[:2]

    # Find the larger dimension
    max_dim = max(height, width)

    # Calculate padding
    pad_height = (max_dim - height) // 2
    pad_width = (max_dim - width) // 2

    # Handle odd dimensions
    pad_height_extra = max_dim - height - 2 * pad_height
    pad_width_extra = max_dim - width - 2 * pad_width

    # Create padding with edge colors
    if len(img.shape) == 3:  # Color image
        # Pad the image
        padded = np.pad(
            img,
            (
                (pad_height, pad_height + pad_height_extra),
                (pad_width, pad_width + pad_width_extra),
                (0, 0),
            ),
            mode="edge",
        )
    else:  # Grayscale image
        padded = np.pad(
            img,
            (
                (pad_height, pad_height + pad_height_extra),
                (pad_width, pad_width + pad_width_extra),
            ),
            mode="edge",
        )

    return padded


def exclude_scaling_box(
    image: np.ndarray,
    bbox: np.ndarray,
    orig_size: tuple,
    processed_size: tuple,
    expansion_factor: float = 1.2,
) -> np.ndarray:
    # Unpack the bounding box
    x_min, y_min, x_max, y_max = map(int, bbox)

    # Calculate scaling factors
    scale_x = processed_size[1] / orig_size[1]  # Width scale
    scale_y = processed_size[0] / orig_size[0]  # Height scale

    # Adjust bounding box coordinates
    x_min = int(x_min * scale_x)
    x_max = int(x_max * scale_x)
    y_min = int(y_min * scale_y)
    y_max = int(y_max * scale_y)

    # Calculate expanded box coordinates
    box_width = x_max - x_min
    box_height = y_max - y_min
    expanded_x_min = max(0, int(x_min - (expansion_factor - 1) * box_width / 2))
    expanded_x_max = min(
        image.shape[1], int(x_max + (expansion_factor - 1) * box_width / 2)
    )
    expanded_y_min = max(0, int(y_min - (expansion_factor - 1) * box_height / 2))
    expanded_y_max = min(
        image.shape[0], int(y_max + (expansion_factor - 1) * box_height / 2)
    )

    # Black out the expanded region
    image[expanded_y_min:expanded_y_max, expanded_x_min:expanded_x_max] = 0

    return image


def resample_contour(contour):
    # Get all the parameters at the start:
    num_points = 1000
    smoothing_factor = 5
    spline_degree = 3  # Typically k=3 for cubic spline

    smoothed_x_sigma = 1
    smoothed_y_sigma = 1

    # Ensure contour has enough points
    if len(contour) < spline_degree + 1:
        raise ValueError(f"Contour must have at least {spline_degree + 1} points, but has {len(contour)} points.")

    contour = contour[:, 0, :]

    tck, _ = splprep([contour[:, 0], contour[:, 1]], s=smoothing_factor)
    u = np.linspace(0, 1, num_points)
    resampled_points = splev(u, tck)

    smoothed_x = gaussian_filter1d(resampled_points[0], sigma=smoothed_x_sigma)
    smoothed_y = gaussian_filter1d(resampled_points[1], sigma=smoothed_y_sigma)

    return np.array([smoothed_x, smoothed_y]).T


def save_dxf_spline(inflated_contours, scaling_factor, height):
    degree = 3
    closed = True

    doc = ezdxf.new(units=0)
    doc.units = ezdxf.units.IN
    doc.header["$INSUNITS"] = ezdxf.units.IN

    msp = doc.modelspace()

    for contour in inflated_contours:
        try:
            resampled_contour = resample_contour(contour)
            points = [
                (x * scaling_factor, (height - y) * scaling_factor)
                for x, y in resampled_contour
            ]
            if len(points) >= 3:
                if np.linalg.norm(np.array(points[0]) - np.array(points[-1])) > 1e-2:
                    points.append(points[0])

                spline = msp.add_spline(points, degree=degree)
                spline.closed = closed
        except ValueError as e:
            print(f"Skipping contour: {e}")

    dxf_filepath = os.path.join("./outputs", "out.dxf")
    doc.saveas(dxf_filepath)
    return dxf_filepath


def extract_outlines(binary_image: np.ndarray) -> np.ndarray:
    """
    Extracts and draws the outlines of masks from a binary image.
    Args:
        binary_image: Grayscale binary image where white represents masks and black is the background.
    Returns:
        Image with outlines drawn.
    """
    # Detect contours from the binary image
    contours, _ = cv2.findContours(
        binary_image, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE
    )

    # Create a blank image to draw contours
    outline_image = np.zeros_like(binary_image)

    # Draw the contours on the blank image
    cv2.drawContours(
        outline_image, contours, -1, (255), thickness=1
    )  # White color for outlines

    return cv2.bitwise_not(outline_image), contours


def shrink_bbox(image: np.ndarray, shrink_factor: float):
    """
    Crops the central portion of the image.
    """
    height, width = image.shape[:2]
    center_x, center_y = width // 2, height // 2

    # Calculate dimensions
    new_width = int(width * shrink_factor)
    new_height = int(height * shrink_factor)

    # Determine the top-left and bottom-right points for cropping
    x1 = max(center_x - new_width // 2, 0)
    y1 = max(center_y - new_height // 2, 0)
    x2 = min(center_x + new_width // 2, width)
    y2 = min(center_y + new_height // 2, height)

    # Crop the image
    cropped_image = image[y1:y2, x1:x2]
    return cropped_image


def to_dxf(contours):
    doc = ezdxf.new()
    msp = doc.modelspace()

    for contour in contours:
        points = [(point[0][0], point[0][1]) for point in contour]
        msp.add_lwpolyline(points, close=True)  # Add a polyline for each contour

    doc.saveas("./outputs/out.dxf")
    return "./outputs/out.dxf"


def smooth_contours(contour):
    epsilon = 0.01 * cv2.arcLength(contour, True)  # Adjust factor (e.g., 0.01)
    return cv2.approxPolyDP(contour, epsilon, True)


def scale_image(image: np.ndarray, scale_factor: float) -> np.ndarray:
    """
    Resize image by scaling both width and height by the same factor.
    """
    if scale_factor <= 0:
        raise ValueError("Scale factor must be positive")

    current_height, current_width = image.shape[:2]

    # Calculate new dimensions
    new_width = int(current_width * scale_factor)
    new_height = int(current_height * scale_factor)

    # Choose interpolation method based on whether we're scaling up or down
    interpolation = cv2.INTER_AREA if scale_factor < 1 else cv2.INTER_CUBIC

    # Resize image
    resized_image = cv2.resize(
        image, (new_width, new_height), interpolation=interpolation
    )

    return resized_image


def detect_reference_square(img) -> np.ndarray:
    box_detector = YOLO("./last.pt")
    res = box_detector.predict(img, conf=0.05)
    del box_detector
    gc.collect()
    return save_one_box(res[0].cpu().boxes.xyxy, res[0].orig_img, save=False), res[
        0
    ].cpu().boxes.xyxy[0]


def resize_img(img: np.ndarray, resize_dim):
    return np.array(Image.fromarray(img).resize(resize_dim))


def predict(image, offset_inches):
    try:
        drawer_img = yolo_detect(image, ["box"])
        shrunked_img = make_square(shrink_bbox(drawer_img, 0.90))
    except:
        raise gr.Error("Unable to DETECT DRAWER, please take another picture with different magnification level!")

    # Detect the scaling reference square
    try:
        reference_obj_img, scaling_box_coords = detect_reference_square(shrunked_img)
    except:
        raise gr.Error("Unable to DETECT REFERENCE BOX, please take another picture with different magnification level!")

    # make the image sqaure so it does not effect the size of objects
    reference_obj_img = make_square(reference_obj_img)
    reference_square_mask = remove_bg(reference_obj_img)

    # make the mask same size as org image
    reference_square_mask = resize_img(
        reference_square_mask, (reference_obj_img.shape[1], reference_obj_img.shape[0])
    )

    scaling_factor = 1.0
    try:
        scaling_factor = calculate_scaling_factor(
            reference_image_path="./Reference_ScalingBox.jpg",
            target_image=reference_square_mask,
            feature_detector="ORB",
        )
    except ZeroDivisionError:
        print("Error calculating scaling factor: Division by zero")
    except Exception as e:
        print(f"Error calculating scaling factor: {e}")

    # Default to a scaling factor of 1.0 if calculation fails or is 0
    if scaling_factor is None or scaling_factor <= 0:
        scaling_factor = 1.0
        print("Using default scaling factor of 1.0 due to calculation error")

    # Save original size before `remove_bg` processing
    orig_size = shrunked_img.shape[:2]
    # Generate foreground mask and save its size
    objects_mask = remove_bg(shrunked_img)

    processed_size = objects_mask.shape[:2]
    # Exclude scaling box region from objects mask
    objects_mask = exclude_scaling_box(
        objects_mask,
        scaling_box_coords,
        orig_size,
        processed_size,
        expansion_factor=1.2,
    )
    objects_mask = resize_img(
        objects_mask, (shrunked_img.shape[1], shrunked_img.shape[0])
    )

    # Ensure offset_inches is valid
    # Calculate pixel dilation amount: (offset_inches / scaling_factor) * 2 + 1
    # We use 1.0 / scaling_factor because scaling_factor is px/inch.
    if scaling_factor > 0:
        # Convert inches to pixels
        offset_pixels = int(offset_inches / scaling_factor * 2) + 1
    else:
        offset_pixels = 1 

    # Dilate mask for offset
    dilated_mask = cv2.dilate(
        objects_mask, np.ones((offset_pixels, offset_pixels), np.uint8)
    )

    Image.fromarray(dilated_mask).save("./outputs/scaled_mask_new.jpg")
    outlines, contours = extract_outlines(dilated_mask)
    shrunked_img_contours = cv2.drawContours(
        shrunked_img, contours, -1, (0, 0, 255), thickness=2
    )
    dxf = save_dxf_spline(contours, scaling_factor, processed_size[0])

    return (
        cv2.cvtColor(shrunked_img_contours, cv2.COLOR_BGR2RGB),
        outlines,
        dxf,
        dilated_mask,
        scaling_factor,
    )


if __name__ == "__main__":
    os.makedirs("./outputs", exist_ok=True)

    # Use gr.Blocks to allow for the structured guideline accordion
    with gr.Blocks(title="Drawer Contouring and DXF Generator") as demo:
        gr.Markdown("<h1 style='text-align: center;'>Drawer Contouring and DXF Generator (YOLO + BiRefNet)</h1>")
        gr.Markdown("Tool for generating scaled manufacturing contours from an input image.")

        # 1. Guidelines Section
        with gr.Accordion(" Tips & User Guidelines", open=False):
            gr.Markdown(GUIDELINE_SETUP)
            gr.Markdown("---")
            gr.Markdown(GUIDELINE_INPUT)
            gr.Markdown("---")
            gr.Markdown(GUIDELINE_OUTPUT)

        # 2. Main Interface
        with gr.Row():
            with gr.Column(scale=1):
                gr.Markdown("## Step 1: Upload an Image")
                input_image = gr.Image(label="1. Input Image", type="numpy")
                gr.Markdown("## Step 2: Set the offset value (Optional) ")
                offset_input = gr.Number(label="2. Offset value for Mask (inches)", value=0.075)
                gr.Markdown("## Step 3: Click the button ")
                submit_button = gr.Button(" Process and Generate DXF", variant="primary")
                

            with gr.Column(scale=2):
                gr.Markdown("## Results")
                scaling_output = gr.Textbox(
                    label="Scaling Factor (pixels/inch)",
                    placeholder="Calculated conversion rate",
                )
                output_image = gr.Image(label="Output Image (Contours Drawn)")
                
                with gr.Row():
                    output_outlines = gr.Image(label="Outlines of Objects")
                    output_mask = gr.Image(label="Final Dilated Mask")
                
                dxf_file = gr.File(label="DXF file (Download)")


        # 3. Examples Section
        gr.Markdown("## Examples ")
        gr.Examples(
            examples=[
                ["./examples/Test20.jpg", 0.075],
                ["./examples/Test21.jpg", 0.075],
                ["./examples/Test22.jpg", 0.075],
                ["./examples/Test23.jpg", 0.075],
            ],
            inputs=[input_image, offset_input],
            outputs=[output_image, output_outlines, dxf_file, output_mask, scaling_output],
            fn=predict,
            cache_examples=False,
            label="Example Images (Click to load and run)",
        )
        
        # Event Handler
        submit_button.click(
            fn=predict,
            inputs=[input_image, offset_input],
            outputs=[output_image, output_outlines, dxf_file, output_mask, scaling_output],
        )

    demo.queue()
    demo.launch(
        server_name="0.0.0.0",
        server_port=7860,
        share=True)