drawing_util.py

import numpy as np

def rgb_to_hex(rgb_color):
    """
    Converts a rgb color to hex color
    Args:
        rgb_color (tuple)
    Returns:
        hex_color (str)
    """
    return "#{:02x}{:02x}{:02x}".format(*rgb_color)

def mask_shifting(mask_data,deleted_mask_id:int,slice_id:int|None = None):
    """
    Shifts the mask when a mask got deleted to restore an order without gaps.

    Args:
        mask_data (np.array): the mask data.
        deleted_mask_id (int): the id of the deleted mask.
        slice_id (int): the id of the slice when the mask is 3d.

    Raises:
          ValueError: if the deleted_mask_id is smaller or equal to 0.
    """
    if deleted_mask_id < 1:
        raise ValueError("deleted_mask_id must be greater than 0")

    mask = mask_data["masks"]
    outline = mask_data["outlines"]

    if mask.ndim == 3:
        if slice_id < 0:
            raise ValueError("slice_id should be non-negative")
        mask_slice = np.take(mask, slice_id, axis=0).astype(np.uint16)
        mask_slice[mask_slice>deleted_mask_id] -= 1
        mask[slice_id, :, :] = mask_slice
    else:
        mask[mask > deleted_mask_id] -= 1

    if outline.ndim == 3:
        if slice_id < 0:
            raise ValueError("slice_id should be non-negative")
        outline_slice = np.take(outline, slice_id, axis=0).astype(np.uint16)
        outline_slice[outline_slice>deleted_mask_id] -= 1
        outline[slice_id, :, :] = outline_slice
    else:
        outline[outline>deleted_mask_id] -= 1

def search_free_id(mask,outline):
    """
    Search in a NumPy array of integers (e.g., [1,1,2,2,3,4,5,5,7,7]) for the first missing number (in this case, 6).
    If no gap is found, return the highest value + 1.
    """

    combined = np.concatenate((mask.ravel(),outline.ravel()))

    unique_vals = np.unique(combined).astype(np.uint16)

    if unique_vals.size == 0:
        return 1

    diffs = np.diff(unique_vals)

    gap_index = np.where(diffs > 1)[0]

    if gap_index.size > 0:
        missed_value = unique_vals[gap_index[0]] + 1
    else:
        missed_value = unique_vals[-1] + 1

    return missed_value

def bresenham_line(start: tuple, end: tuple):
    """
    Calculates all pixel coordinates along a line from start to end using the Bresenham algorithm.
    """
    x0, y0 = int(round(start[0])), int(round(start[1]))
    x1, y1 = int(round(end[0])), int(round(end[1]))
    pixels = []

    dx = abs(x1 - x0)
    dy = abs(y1 - y0)
    sx = 1 if x0 < x1 else -1
    sy = 1 if y0 < y1 else -1
    err = dx - dy

    while True:
        pixels.append((x0, y0))
        if x0 == x1 and y0 == y1:
            break
        e2 = 2 * err
        if e2 > -dy:
            err -= dy
            x0 += sx
        if e2 < dx:
            err += dx
            y0 += sy

    return pixels

def trace_contour(binary_mask):
    """
    This method traces the contours of the inputted mask and returns the outline
    Attributes:
        binary_mask (np.array): the binary mask.

    Returns:
        contour: the outline of the cells in the mask

    """
    y_indices, x_indices = np.where(binary_mask)
    start_idx = np.lexsort((x_indices, y_indices))[0]
    start_x, start_y = x_indices[start_idx], y_indices[start_idx]

    directions = [(0, -1), (1, -1), (1, 0), (1, 1), (0, 1), (-1, 1), (-1, 0), (-1, -1)]
    contour = []
    current = (start_x, start_y)
    prev = (start_x - 1, start_y)

    while True:
        contour.append(current)
        dir_idx = directions.index((prev[0] - current[0], prev[1] - current[1]))
        for i in range(8):
            next_dir = (dir_idx + 1 + i) % 8
            dx, dy = directions[next_dir]
            nx, ny = current[0] + dx, current[1] + dy
            if 0 <= ny < binary_mask.shape[0] and 0 <= nx < binary_mask.shape[1]:
                if binary_mask[ny, nx]:
                    prev = current
                    current = (nx, ny)
                    break
        if current == (start_x, start_y):
            break
    return contour

def fill_polygon_from_outline(contour, mask_shape):
    """
    Fills the given polygon in contour with the scanline technique.
    Attributes:
        contour (np.array): the contour
        mask_shape (np.array): the mask dimensions

    Returns:
        mask: the adapted mask with newly filled cells
    """
    mask = np.zeros(mask_shape, dtype=np.uint16)
    if not contour:
        return mask

    edges = []
    num_vertices = len(contour)
    for i in range(num_vertices):
        x0, y0 = contour[i]
        x1, y1 = contour[(i + 1) % num_vertices]
        if y0 == y1:
            continue  # Skip horizontal edges
        if y0 > y1:
            x0, y0, x1, y1 = x1, y1, x0, y0  # Ensure y0 < y1
        slope = (x1 - x0) / (y1 - y0) if (y1 - y0) != 0 else 0
        edges.append((y0, y1, x0, slope))

    if not edges:
        return mask

    global_edge_table = sorted(edges, key=lambda e: (e[0], e[2]))
    scan_line = min(e[0] for e in global_edge_table)
    active_edges = []

    while global_edge_table or active_edges:
        # Add edges starting at the current scan line
        while global_edge_table and global_edge_table[0][0] == scan_line:
            e = global_edge_table.pop(0)
            active_edges.append([e[1], e[2], e[3]])

        active_edges.sort(key=lambda e: e[1])

        # Fill between pairs of edges
        for i in range(0, len(active_edges), 2):
            if i + 1 >= len(active_edges):
                break
            e1 = active_edges[i]
            e2 = active_edges[i + 1]
            x_start = int(np.ceil(e1[1]))
            x_end = int(np.floor(e2[1]))
            for x in range(x_start, x_end + 1):
                if 0 <= x < mask_shape[1] and 0 <= scan_line < mask_shape[0]:
                    mask[scan_line, x] = 1

        scan_line += 1
        active_edges = [e for e in active_edges if e[0] > scan_line]
        for e in active_edges:
            e[1] += e[2]

    return mask

def find_border_pixels(mask, outline, cell_id,):
    """
    Finds edge pixels in a given mask matrix, considering the outline and only searching for pixels that match the specified cell_id.
    Edge pixels are those whose neighboring pixels have a different ID or are marked as edges in the outline.
    Attributes:
        mask: A 2D mask representing the ID of each pixel in the area.
        outline: A 2D mask marking the edge pixels (typically with a value of 100).
        cell_id: The ID of the cell whose edge pixels should be found.

    Returns:
        border_pixels: all neighbours of the cell
    """
    border_pixels = []

    rows, cols = mask.shape
    for y in range(rows):
        for x in range(cols):
            # ID of current pixel
            current_id = mask[y, x]

            if current_id != cell_id:
                continue

            # neighboring positions
            neighbors = [
                (y - 1, x), (y + 1, x),  # up, down
                (y, x - 1), (y, x + 1)  # left, right
            ]

            is_border_pixel = False

            for ny, nx in neighbors:
                if 0 <= ny < rows and 0 <= nx < cols:
                    if mask[ny, nx] != current_id and outline[ny, nx] != current_id:
                        is_border_pixel = True
                        break
            if is_border_pixel:
                border_pixels.append((y,x))

    return border_pixels