466 lines
17 KiB
Python
466 lines
17 KiB
Python
"""Index generation: NDVI and GCC from S2/S3/fusion GeoTIFFs."""
|
||
import json
|
||
import numpy as np
|
||
import rasterio
|
||
from rasterio.warp import transform as transform_coords
|
||
from pathlib import Path
|
||
from datetime import datetime
|
||
|
||
from preselection import _sample_3x3
|
||
|
||
RED_BAND = 3
|
||
NIR_BAND = 4
|
||
BLUE_BAND = 1
|
||
GREEN_BAND = 2
|
||
|
||
|
||
def _calculate_and_write_ndvi(input_file, output_file):
|
||
with rasterio.open(input_file) as src:
|
||
red = src.read(RED_BAND).astype(np.float32)
|
||
nir = src.read(NIR_BAND).astype(np.float32)
|
||
|
||
mask = (red > 0) & (nir > 0)
|
||
ndvi = np.zeros_like(red, dtype=np.float32)
|
||
ndvi[mask] = (nir[mask] - red[mask]) / (nir[mask] + red[mask])
|
||
|
||
profile = src.profile.copy()
|
||
profile.update(
|
||
{
|
||
"count": 1,
|
||
"dtype": "float32",
|
||
"nodata": 0,
|
||
"compress": "lzw",
|
||
}
|
||
)
|
||
|
||
with rasterio.open(output_file, "w", **profile) as dst:
|
||
dst.write(ndvi, 1)
|
||
dst.set_band_description(1, "NDVI")
|
||
|
||
|
||
def _get_ndvi_value(ndvi_file, site_position):
|
||
try:
|
||
with rasterio.open(ndvi_file) as src:
|
||
lon, lat = site_position[1], site_position[0]
|
||
x, y = transform_coords("EPSG:4326", src.crs, [lon], [lat])
|
||
|
||
# Check if point is within bounds
|
||
if not (
|
||
src.bounds.left <= x[0] <= src.bounds.right
|
||
and src.bounds.bottom <= y[0] <= src.bounds.top
|
||
):
|
||
return None # Point is outside raster bounds
|
||
|
||
samples = list(src.sample([(x[0], y[0])]))
|
||
if samples:
|
||
value = float(samples[0][0])
|
||
# Check if it's actually nodata (using raster's nodata value)
|
||
if src.nodata is not None and value == src.nodata:
|
||
return None # This is nodata, not a valid 0 value
|
||
if np.isnan(value):
|
||
return None # NaN is invalid
|
||
# 0 is a valid NDVI value (no vegetation), so return it
|
||
return value
|
||
except Exception as e:
|
||
print(f"Error sampling {ndvi_file.name}: {e}")
|
||
pass
|
||
return None
|
||
|
||
|
||
def _create_timeseries_for_dir(input_dir, output_dir, site_position, source_name, pattern="*.geotiff"):
|
||
print(f"[NDVI-{source_name}] Creating timeseries.json...")
|
||
timeseries = []
|
||
|
||
for input_file in sorted(input_dir.glob(pattern)):
|
||
if "DIST_CLOUD" in input_file.name:
|
||
continue
|
||
|
||
filename = input_file.name
|
||
parts = filename.replace(".geotiff", "").split("_")
|
||
date_str = None
|
||
|
||
for part in parts:
|
||
if len(part) == 8 and part.isdigit():
|
||
date_str = part
|
||
break
|
||
|
||
if date_str:
|
||
try:
|
||
date = datetime.strptime(date_str, "%Y%m%d").isoformat()
|
||
except ValueError:
|
||
date = date_str
|
||
else:
|
||
date_str = parts[0]
|
||
date = date_str
|
||
print(
|
||
f"[NDVI-{source_name}] Warning: Could not extract date from {filename}, using '{date_str}'"
|
||
)
|
||
|
||
ndvi_value, band_means = _sample_3x3(input_file, site_position)
|
||
blue_mean = band_means.get("b02") if band_means else None
|
||
if ndvi_value is None:
|
||
print(
|
||
f"[NDVI-{source_name}] Warning: Could not sample {filename} (outside bounds or nodata)"
|
||
)
|
||
|
||
entry = {"date": date, "filename": filename, "ndvi": ndvi_value}
|
||
if blue_mean is not None:
|
||
entry["blue"] = blue_mean
|
||
timeseries.append(entry)
|
||
|
||
timeseries.sort(key=lambda x: x["date"])
|
||
output_dir.mkdir(parents=True, exist_ok=True)
|
||
timeseries_file = output_dir / "timeseries.json"
|
||
with open(timeseries_file, "w") as f:
|
||
json.dump(timeseries, f, indent=2)
|
||
|
||
print(f"[NDVI-{source_name}] Saved: {timeseries_file} ({len(timeseries)} entries)")
|
||
|
||
|
||
def _process_ndvi_files(
|
||
input_dir, output_dir, source_name, pattern="*.geotiff", output_namer=None
|
||
):
|
||
output_dir.mkdir(parents=True, exist_ok=True)
|
||
print(f"[NDVI-{source_name}] Processing {input_dir}...")
|
||
|
||
geotiff_files = sorted(input_dir.glob(pattern))
|
||
if not geotiff_files:
|
||
print(f"[NDVI-{source_name}] No files found")
|
||
return
|
||
|
||
for geotiff_file in geotiff_files:
|
||
# Skip DIST_CLOUD files silently (single-band distance-to-clouds, not suitable for NDVI)
|
||
if "DIST_CLOUD" in geotiff_file.name:
|
||
continue
|
||
|
||
# Check if file has enough bands (need at least 4 for RED and NIR)
|
||
try:
|
||
with rasterio.open(geotiff_file) as src:
|
||
if src.count < 4:
|
||
print(
|
||
f"[NDVI-{source_name}] Skipping {geotiff_file.name} (only {src.count} band(s), need 4+)"
|
||
)
|
||
continue
|
||
except Exception as e:
|
||
print(
|
||
f"[NDVI-{source_name}] Skipping {geotiff_file.name} (error reading: {e})"
|
||
)
|
||
continue
|
||
|
||
output_file = output_dir / (
|
||
output_namer(geotiff_file) if output_namer else geotiff_file.name
|
||
)
|
||
|
||
_calculate_and_write_ndvi(geotiff_file, output_file)
|
||
print(f"[NDVI-{source_name}] Saved: {output_file}")
|
||
|
||
|
||
def generate_ndvi_raw(season, site_position, site_name):
|
||
# No longer creating NDVI GeoTIFF files, only timeseries
|
||
pass
|
||
|
||
|
||
def _get_output_name_prepared(geotiff_file):
|
||
if geotiff_file.suffix == ".tif":
|
||
if "REFL" in geotiff_file.stem:
|
||
# For S2: S2A_MSIL2A_20240101_REFL -> date is at index [2]
|
||
# For S3: composite_20240101.tif -> date is at index [1] after removing .tif
|
||
parts = geotiff_file.stem.split("_")
|
||
if len(parts) >= 3 and parts[0].startswith("S2"):
|
||
# S2 format: S2A_MSIL2A_YYYYMMDD_REFL
|
||
date_str = parts[2]
|
||
elif len(parts) >= 2 and parts[0] == "composite":
|
||
# S3 format: composite_YYYYMMDD
|
||
date_str = parts[1]
|
||
else:
|
||
# Fallback: try index [1] for other formats
|
||
date_str = parts[1] if len(parts) > 1 else parts[0]
|
||
return f"{date_str}_ndvi.geotiff"
|
||
return geotiff_file.name.replace(".tif", ".geotiff")
|
||
return geotiff_file.name
|
||
|
||
|
||
def _fusion_namer(f):
|
||
date_str = f.stem.split("_")[1]
|
||
return f"{date_str}_ndvi.geotiff"
|
||
|
||
|
||
def generate_ndvi_post_process(season, site_position, site_name):
|
||
# No longer creating NDVI GeoTIFF files, only timeseries
|
||
pass
|
||
|
||
|
||
def create_ndvi_timeseries_post_process(season, site_position, site_name):
|
||
for strategy in ["aggressive", "nonaggressive"]:
|
||
for sigma in [20, 30]:
|
||
processed_dir = f"processed_{strategy}_sigma{sigma}"
|
||
for source in ["s2", "s3"]:
|
||
input_dir = Path(f"data/{site_name}/{season}/{processed_dir}/{source}/")
|
||
output_dir = Path(f"data/{site_name}/{season}/{processed_dir}/ndvi/{source}/")
|
||
_create_timeseries_for_dir(
|
||
input_dir, output_dir, site_position, f"POST-PROCESS-{source.upper()}-{strategy}-σ{sigma}"
|
||
)
|
||
input_dir = Path(f"data/{site_name}/{season}/{processed_dir}/fusion/")
|
||
output_dir = Path(f"data/{site_name}/{season}/{processed_dir}/ndvi/fusion/")
|
||
_create_timeseries_for_dir(input_dir, output_dir, site_position, f"POST-PROCESS-FUSION-{strategy}-σ{sigma}")
|
||
|
||
|
||
def _calculate_and_write_gcc(input_file, output_file):
|
||
with rasterio.open(input_file) as src:
|
||
blue = src.read(BLUE_BAND).astype(np.float32)
|
||
green = src.read(GREEN_BAND).astype(np.float32)
|
||
red = src.read(RED_BAND).astype(np.float32)
|
||
|
||
total = red + green + blue
|
||
mask = total > 0
|
||
gcc = np.zeros_like(green, dtype=np.float32)
|
||
gcc[mask] = green[mask] / total[mask]
|
||
|
||
profile = src.profile.copy()
|
||
profile.update(
|
||
{
|
||
"count": 1,
|
||
"dtype": "float32",
|
||
"nodata": 0,
|
||
"compress": "lzw",
|
||
}
|
||
)
|
||
|
||
with rasterio.open(output_file, "w", **profile) as dst:
|
||
dst.write(gcc, 1)
|
||
dst.set_band_description(1, "GCC")
|
||
|
||
|
||
def _get_gcc_value(gcc_file, site_position):
|
||
try:
|
||
with rasterio.open(gcc_file) as src:
|
||
lon, lat = site_position[1], site_position[0]
|
||
x, y = transform_coords("EPSG:4326", src.crs, [lon], [lat])
|
||
|
||
if not (
|
||
src.bounds.left <= x[0] <= src.bounds.right
|
||
and src.bounds.bottom <= y[0] <= src.bounds.top
|
||
):
|
||
return None
|
||
|
||
samples = list(src.sample([(x[0], y[0])]))
|
||
if samples:
|
||
value = float(samples[0][0])
|
||
if src.nodata is not None and value == src.nodata:
|
||
return None
|
||
if np.isnan(value):
|
||
return None
|
||
return value
|
||
except Exception as e:
|
||
print(f"Error sampling {gcc_file.name}: {e}")
|
||
pass
|
||
return None
|
||
|
||
|
||
def _get_gcc_from_original(input_file, site_position):
|
||
"""Calculate GCC directly from original file without creating GeoTIFF."""
|
||
try:
|
||
with rasterio.open(input_file) as src:
|
||
if src.count < 3:
|
||
return None
|
||
|
||
blue = src.read(BLUE_BAND).astype(np.float32)
|
||
green = src.read(GREEN_BAND).astype(np.float32)
|
||
red = src.read(RED_BAND).astype(np.float32)
|
||
|
||
lon, lat = site_position[1], site_position[0]
|
||
x, y = transform_coords("EPSG:4326", src.crs, [lon], [lat])
|
||
|
||
if not (
|
||
src.bounds.left <= x[0] <= src.bounds.right
|
||
and src.bounds.bottom <= y[0] <= src.bounds.top
|
||
):
|
||
return None
|
||
|
||
row, col = src.index(x[0], y[0])
|
||
if row < 0 or row >= src.height or col < 0 or col >= src.width:
|
||
return None
|
||
|
||
# Extract 3x3 window with boundary handling
|
||
r0, r1 = max(0, row - 1), min(src.height, row + 2)
|
||
c0, c1 = max(0, col - 1), min(src.width, col + 2)
|
||
blue_window = blue[r0:r1, c0:c1]
|
||
green_window = green[r0:r1, c0:c1]
|
||
red_window = red[r0:r1, c0:c1]
|
||
|
||
# Calculate GCC for each pixel in window
|
||
total = red_window + green_window + blue_window
|
||
mask = (total > 0) & ~np.isnan(total) & (blue_window >= 0) & (green_window >= 0) & (red_window >= 0)
|
||
if not np.any(mask):
|
||
negative_pixels = np.sum((blue_window < 0) | (green_window < 0) | (red_window < 0))
|
||
if negative_pixels > 0:
|
||
print(f"Warning: {input_file.name} excluded - all pixels have negative band values ({negative_pixels} negative pixels in window)")
|
||
return None
|
||
|
||
gcc_window = np.zeros_like(green_window, dtype=np.float32)
|
||
gcc_window[mask] = green_window[mask] / total[mask]
|
||
|
||
# Return mean of valid GCC values
|
||
valid_gcc = gcc_window[mask]
|
||
return float(np.mean(valid_gcc)) if len(valid_gcc) > 0 else None
|
||
except Exception as e:
|
||
return None
|
||
|
||
|
||
def _create_gcc_timeseries_for_dir(input_dir, output_dir, site_position, source_name, pattern="*.geotiff"):
|
||
print(f"[GCC-{source_name}] Creating timeseries.json...")
|
||
timeseries = []
|
||
|
||
for input_file in sorted(input_dir.glob(pattern)):
|
||
if "DIST_CLOUD" in input_file.name:
|
||
continue
|
||
|
||
filename = input_file.name
|
||
parts = filename.replace(".geotiff", "").split("_")
|
||
date_str = None
|
||
|
||
for part in parts:
|
||
if len(part) == 8 and part.isdigit():
|
||
date_str = part
|
||
break
|
||
|
||
if date_str:
|
||
try:
|
||
date = datetime.strptime(date_str, "%Y%m%d").isoformat()
|
||
except ValueError:
|
||
date = date_str
|
||
else:
|
||
date_str = parts[0]
|
||
date = date_str
|
||
print(
|
||
f"[GCC-{source_name}] Warning: Could not extract date from {filename}, using '{date_str}'"
|
||
)
|
||
|
||
gcc_value = _get_gcc_from_original(input_file, site_position)
|
||
if gcc_value is None:
|
||
print(
|
||
f"[GCC-{source_name}] Warning: Could not sample {filename} (outside bounds or nodata)"
|
||
)
|
||
|
||
timeseries.append({"date": date, "filename": filename, "greenness_index": gcc_value})
|
||
|
||
timeseries.sort(key=lambda x: x["date"])
|
||
output_dir.mkdir(parents=True, exist_ok=True)
|
||
timeseries_file = output_dir / "timeseries.json"
|
||
with open(timeseries_file, "w") as f:
|
||
json.dump(timeseries, f, indent=2)
|
||
|
||
print(f"[GCC-{source_name}] Saved: {timeseries_file} ({len(timeseries)} entries)")
|
||
|
||
|
||
def _process_gcc_files(
|
||
input_dir, output_dir, source_name, pattern="*.geotiff", output_namer=None
|
||
):
|
||
output_dir.mkdir(parents=True, exist_ok=True)
|
||
print(f"[GCC-{source_name}] Processing {input_dir}...")
|
||
|
||
geotiff_files = sorted(input_dir.glob(pattern))
|
||
if not geotiff_files:
|
||
print(f"[GCC-{source_name}] No files found")
|
||
return
|
||
|
||
for geotiff_file in geotiff_files:
|
||
if "DIST_CLOUD" in geotiff_file.name:
|
||
continue
|
||
|
||
try:
|
||
with rasterio.open(geotiff_file) as src:
|
||
if src.count < 3:
|
||
print(
|
||
f"[GCC-{source_name}] Skipping {geotiff_file.name} (only {src.count} band(s), need 3+)"
|
||
)
|
||
continue
|
||
except Exception as e:
|
||
print(
|
||
f"[GCC-{source_name}] Skipping {geotiff_file.name} (error reading: {e})"
|
||
)
|
||
continue
|
||
|
||
output_file = output_dir / (
|
||
output_namer(geotiff_file) if output_namer else geotiff_file.name
|
||
)
|
||
|
||
_calculate_and_write_gcc(geotiff_file, output_file)
|
||
print(f"[GCC-{source_name}] Saved: {output_file}")
|
||
|
||
|
||
def generate_gcc_post_process(season, site_position, site_name):
|
||
# No longer creating GCC GeoTIFF files, only timeseries
|
||
pass
|
||
|
||
|
||
def create_gcc_timeseries_post_process(season, site_position, site_name):
|
||
for strategy in ["aggressive", "nonaggressive"]:
|
||
for sigma in [20, 30]:
|
||
processed_dir = f"processed_{strategy}_sigma{sigma}"
|
||
for source in ["s2", "s3"]:
|
||
input_dir = Path(f"data/{site_name}/{season}/{processed_dir}/{source}/")
|
||
output_dir = Path(f"data/{site_name}/{season}/{processed_dir}/gcc/{source}/")
|
||
_create_gcc_timeseries_for_dir(
|
||
input_dir, output_dir, site_position, f"POST-PROCESS-{source.upper()}-{strategy}-σ{sigma}"
|
||
)
|
||
input_dir = Path(f"data/{site_name}/{season}/{processed_dir}/fusion/")
|
||
output_dir = Path(f"data/{site_name}/{season}/{processed_dir}/gcc/fusion/")
|
||
_create_gcc_timeseries_for_dir(input_dir, output_dir, site_position, f"POST-PROCESS-FUSION-{strategy}-σ{sigma}")
|
||
|
||
|
||
def _get_bands_from_original(input_file, site_position):
|
||
"""Extract mean B02, B03, B04, B8A from 3x3 window at site. Returns dict or None."""
|
||
try:
|
||
with rasterio.open(input_file) as src:
|
||
if src.count < 4:
|
||
return None
|
||
lon, lat = site_position[1], site_position[0]
|
||
x, y = transform_coords("EPSG:4326", src.crs, [lon], [lat])
|
||
if not (
|
||
src.bounds.left <= x[0] <= src.bounds.right
|
||
and src.bounds.bottom <= y[0] <= src.bounds.top
|
||
):
|
||
return None
|
||
row, col = src.index(x[0], y[0])
|
||
r0, r1 = max(0, row - 1), min(src.height, row + 2)
|
||
c0, c1 = max(0, col - 1), min(src.width, col + 2)
|
||
bands = [src.read(i + 1, window=((r0, r1), (c0, c1))).astype(np.float32) for i in range(4)]
|
||
mask = ~np.any([np.isnan(b) for b in bands], axis=0)
|
||
mask &= np.all([b > 0 for b in bands], axis=0)
|
||
if not np.any(mask):
|
||
return None
|
||
return {
|
||
"b02": float(np.mean(bands[0][mask])),
|
||
"b03": float(np.mean(bands[1][mask])),
|
||
"b04": float(np.mean(bands[2][mask])),
|
||
"b8a": float(np.mean(bands[3][mask])),
|
||
}
|
||
except Exception:
|
||
return None
|
||
|
||
|
||
def _create_s2_bands_timeseries_for_dir(input_dir, output_dir, site_position):
|
||
print(f"[S2-BANDS] Creating timeseries.json...")
|
||
timeseries = []
|
||
for f in sorted(input_dir.glob("*.geotiff")):
|
||
date_str = f.stem.split("_")[0]
|
||
if len(date_str) != 8 or not date_str.isdigit():
|
||
continue
|
||
date = datetime.strptime(date_str, "%Y%m%d").isoformat()
|
||
bands = _get_bands_from_original(f, site_position)
|
||
timeseries.append({"date": date, "filename": f.name, **(bands or {})})
|
||
timeseries.sort(key=lambda x: x["date"])
|
||
output_dir.mkdir(parents=True, exist_ok=True)
|
||
(output_dir / "timeseries.json").write_text(json.dumps(timeseries, indent=2))
|
||
print(f"[S2-BANDS] Saved: {output_dir / 'timeseries.json'} ({len(timeseries)} entries)")
|
||
|
||
|
||
def create_s2_bands_timeseries_post_process(season, site_position, site_name):
|
||
for strategy in ["aggressive", "nonaggressive"]:
|
||
for sigma in [20, 30]:
|
||
processed_dir = f"processed_{strategy}_sigma{sigma}"
|
||
input_dir = Path(f"data/{site_name}/{season}/{processed_dir}/s2/")
|
||
output_dir = Path(f"data/{site_name}/{season}/{processed_dir}/s2_bands/")
|
||
if input_dir.exists():
|
||
_create_s2_bands_timeseries_for_dir(input_dir, output_dir, site_position)
|