File size: 2,525 Bytes
2501dd9 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 | """Visual enhancement modes for the displayed cutout (apoyo al revisor).
Same transforms and parameters as the static review site (docs/app.js), ported
to numpy/skimage. These are cosmetic, applied to the 8-bit RGB only — they do
NOT touch the image the model sees (that stays the raw Lupton composite).
"""
from __future__ import annotations
import numpy as np
from skimage.exposure import equalize_adapthist
from skimage.filters import unsharp_mask
MODES = ["Original", "Percentil", "Unsharp", "Asinh", "CLAHE", "Estructuras"]
def percentile_stretch(rgb: np.ndarray, p_lo: float = 1.0, p_hi: float = 99.5) -> np.ndarray:
out = rgb.astype(np.float32)
for c in range(3):
lo, hi = np.percentile(out[..., c], [p_lo, p_hi])
rng = (hi - lo) or 1.0
out[..., c] = np.clip((out[..., c] - lo) / rng * 255.0, 0, 255)
return out.astype(np.uint8)
def unsharp(rgb: np.ndarray, radius: float = 6.0, amount: float = 1.5) -> np.ndarray:
f = rgb.astype(np.float32) / 255.0
out = unsharp_mask(f, radius=radius, amount=amount, channel_axis=-1)
return (np.clip(out, 0, 1) * 255).astype(np.uint8)
def asinh_stretch(rgb: np.ndarray, a: float = 1.5) -> np.ndarray:
f = rgb.astype(np.float32) / 255.0
out = np.arcsinh(a * f) / np.arcsinh(a)
return (np.clip(out, 0, 1) * 255).astype(np.uint8)
def clahe(rgb: np.ndarray, clip_limit: float = 0.01) -> np.ndarray:
f = rgb.astype(np.float32) / 255.0
out = np.zeros_like(f)
for c in range(3):
out[..., c] = equalize_adapthist(f[..., c], clip_limit=clip_limit)
return (np.clip(out, 0, 1) * 255).astype(np.uint8)
def estructuras(rgb: np.ndarray, amount: float = 1.5) -> np.ndarray:
x = percentile_stretch(rgb, 1.0, 99.5)
x = unsharp(x, 6.0, amount)
x = asinh_stretch(x, amount * 0.5)
return x
def to_gray(rgb: np.ndarray) -> np.ndarray:
lum = (0.2126 * rgb[..., 0] + 0.7152 * rgb[..., 1] + 0.0722 * rgb[..., 2]).astype(np.uint8)
return np.dstack([lum, lum, lum])
def apply(rgb: np.ndarray, mode: str = "Original", gray: bool = False) -> np.ndarray:
"""Apply the chosen enhancement mode (and optional grayscale) to an RGB."""
if mode == "Percentil":
rgb = percentile_stretch(rgb)
elif mode == "Unsharp":
rgb = unsharp(rgb)
elif mode == "Asinh":
rgb = asinh_stretch(rgb)
elif mode == "CLAHE":
rgb = clahe(rgb)
elif mode == "Estructuras":
rgb = estructuras(rgb)
if gray:
rgb = to_gray(rgb)
return rgb
|