modosaic.depth.visualization

modosaic.depth.visualization

DepthVisualizationService

Convert depth arrays into image visualizations.

depth_to_heatmap staticmethod

depth_to_heatmap(depth, size=None, colormap=cv2.COLORMAP_TURBO)

Convert a depth map to a color heatmap.

Parameters:

Name	Type	Description	Default
depth	ndarray	Depth map.	required
size	tuple[int, int] | None	Optional output (width, height).	None
colormap	int	OpenCV color-map identifier.	COLORMAP_TURBO

Returns:

Type	Description
Image	RGB PIL heatmap image.

Source code in modosaic/depth/visualization.py

@staticmethod
def depth_to_heatmap(
        depth: np.ndarray,
        size: tuple[int, int] | None = None,
        colormap: int = cv2.COLORMAP_TURBO,
) -> Image.Image:
    """Convert a depth map to a color heatmap.

    Args:
        depth: Depth map.
        size: Optional output `(width, height)`.
        colormap: OpenCV color-map identifier.

    Returns:
        RGB PIL heatmap image.
    """
    depth_uint8, _, _ = DepthVisualizationService.normalize_to_uint8(depth)
    if size is not None:
        depth_uint8 = cv2.resize(depth_uint8, size, interpolation=cv2.INTER_CUBIC)

    heatmap_bgr = cv2.applyColorMap(depth_uint8, colormap)
    heatmap_rgb = cv2.cvtColor(heatmap_bgr, cv2.COLOR_BGR2RGB)
    return Image.fromarray(heatmap_rgb)

depth_to_image staticmethod

depth_to_image(depth, size=None)

Convert a depth map to a grayscale PIL image.

Parameters:

Name	Type	Description	Default
depth	ndarray	Depth map.	required
size	tuple[int, int] | None	Optional output (width, height).	None

Returns:

Type	Description
Image	PIL image visualization.

Source code in modosaic/depth/visualization.py

@staticmethod
def depth_to_image(
        depth: np.ndarray,
        size: tuple[int, int] | None = None,
) -> Image.Image:
    """Convert a depth map to a grayscale PIL image.

    Args:
        depth: Depth map.
        size: Optional output `(width, height)`.

    Returns:
        PIL image visualization.
    """
    depth_uint8, _, _ = DepthVisualizationService.normalize_to_uint8(depth)
    depth_image = Image.fromarray(depth_uint8)
    if size is not None:
        depth_image = depth_image.resize(size, Image.Resampling.BICUBIC)
    return depth_image

normalize_to_uint8 staticmethod

normalize_to_uint8(depth, eps=1e-08)

Normalize a depth map to an 8-bit grayscale array.

Parameters:

Name	Type	Description	Default
depth	ndarray	Depth map.	required
eps	float	Minimum denominator used to avoid division by zero.	1e-08

Returns:

Type	Description
tuple[ndarray, float, float]	Normalized uint8 array, original minimum, and original maximum.

Source code in modosaic/depth/visualization.py

@staticmethod
def normalize_to_uint8(
        depth: np.ndarray,
        eps: float = 1e-8,
) -> tuple[np.ndarray, float, float]:
    """Normalize a depth map to an 8-bit grayscale array.

    Args:
        depth: Depth map.
        eps: Minimum denominator used to avoid division by zero.

    Returns:
        Normalized uint8 array, original minimum, and original maximum.
    """
    depth = depth.astype(np.float32)
    d_min = float(np.nanmin(depth))
    d_max = float(np.nanmax(depth))
    denom = (d_max - d_min) if (d_max - d_min) > eps else eps
    depth_uint8 = ((depth - d_min) / denom * 255.0).clip(0, 255).astype(np.uint8)
    return depth_uint8, d_min, d_max

overlay_depth staticmethod

overlay_depth(image, depth, alpha=0.55)

Blend a depth preview over an RGB image.

Parameters:

Name	Type	Description	Default
image	Image	Source image.	required
depth	ndarray	Depth map.	required
alpha	float	Blend factor for the depth image.	0.55

Returns:

Type	Description
Image	Blended RGB image.

Source code in modosaic/depth/visualization.py

@staticmethod
def overlay_depth(
        image: Image.Image,
        depth: np.ndarray,
        alpha: float = 0.55,
) -> Image.Image:
    """Blend a depth preview over an RGB image.

    Args:
        image: Source image.
        depth: Depth map.
        alpha: Blend factor for the depth image.

    Returns:
        Blended RGB image.
    """
    base = image.convert("RGB")
    depth_image = DepthVisualizationService.depth_to_image(
        depth,
        size=base.size,
    ).convert("RGB")
    return Image.blend(base, depth_image, alpha=float(np.clip(alpha, 0.0, 1.0)))