add cloud masking for S2 SR images

limiao766 · May 8, 2020 · 25a3f9e · 25a3f9e
1 parent a5adc95
commit 25a3f9e
Show file tree

Hide file tree

Showing 3 changed files with 34 additions and 17 deletions.
diff --git a/.spyproject/workspace.ini b/.spyproject/workspace.ini
@@ -6,5 +6,5 @@ save_non_project_files = False
 
 [main]
 version = 0.1.0
-recent_files = ['C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\calibration.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\clouds.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\stats.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\processing.py']
+recent_files = ['C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\calibration.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\clouds.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\stats.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\processing.py', 'C:\\Users\\mevans\\OneDrive - Defenders of Wildlife\\repos\\Satellite_ComputerVision\\utils\\model.py']
 
diff --git a/utils/clouds.py b/utils/clouds.py
@@ -29,6 +29,23 @@ def rescale(img, exp, thresholds):
     #return img.subtract(thresholds[0]).divide(thresholds[1]-thresholds[0])
     return img.expression(exp, {'img': img}).subtract(thresholds[0]).divide(thresholds[1] - thresholds[0])
 
+def maskS2SR(img):
+    """
+    Apply built in masks to Sentinel-2 surface reflectance imagery
+    Parameters:
+        img (ee.Image): Sentinel-2 level 2A surface reflectange image
+    Returns:
+        ee.Image: masked image
+    """
+    scored = basicQA(img)
+    maskBand = img.select('SCL')
+    cloudMask = maskBand.neq(8).And(maskBand.neq(9))
+    waterMask = maskBand.neq(6)
+    cirrusMask = maskBand.neq(10)
+    snowMask = maskBand.neq(11)
+    darkMask = maskBand.neq(2).And(maskBand.neq(3))
+    return scored.updateMask(cloudMask.And(waterMask).And(cirrusMask).And(snowMask).And(darkMask))
+
 def waterScore(img):
     """ 
     Calculate a water likelihood score [0, 1]

diff --git a/utils/model.py b/utils/model.py
@@ -73,26 +73,26 @@ def get_model(depth, optim, mets):
     Returns:
         tf.keras.model: compiled U-Net model
     """
-	inputs = layers.Input(shape=[None, None, len(BANDS)]) # 256
-	encoder0_pool, encoder0 = encoder_block(inputs, 32) # 128
-	encoder1_pool, encoder1 = encoder_block(encoder0_pool, 64) # 64
-	encoder2_pool, encoder2 = encoder_block(encoder1_pool, 128) # 32
-	encoder3_pool, encoder3 = encoder_block(encoder2_pool, 256) # 16
-	encoder4_pool, encoder4 = encoder_block(encoder3_pool, 512) # 8
-	center = conv_block(encoder4_pool, 1024) # center
-	decoder4 = decoder_block(center, encoder4, 512) # 16
-	decoder3 = decoder_block(decoder4, encoder3, 256) # 32
-	decoder2 = decoder_block(decoder3, encoder2, 128) # 64
-	decoder1 = decoder_block(decoder2, encoder1, 64) # 128
-	decoder0 = decoder_block(decoder1, encoder0, 32) # 256
-	outputs = layers.Conv2D(1, (1, 1), activation='sigmoid')(decoder0)
+    inputs = layers.Input(shape=[None, None, len(BANDS)]) # 256
+    encoder0_pool, encoder0 = encoder_block(inputs, 32) # 128
+    encoder1_pool, encoder1 = encoder_block(encoder0_pool, 64) # 64
+    encoder2_pool, encoder2 = encoder_block(encoder1_pool, 128) # 32
+    encoder3_pool, encoder3 = encoder_block(encoder2_pool, 256) # 16
+    encoder4_pool, encoder4 = encoder_block(encoder3_pool, 512) # 8
+    center = conv_block(encoder4_pool, 1024) # center
+    decoder4 = decoder_block(center, encoder4, 512) # 16
+    decoder3 = decoder_block(decoder4, encoder3, 256) # 32
+    decoder2 = decoder_block(decoder3, encoder2, 128) # 64
+    decoder1 = decoder_block(decoder2, encoder1, 64) # 128
+    decoder0 = decoder_block(decoder1, encoder0, 32) # 256
+    outputs = layers.Conv2D(1, (1, 1), activation='sigmoid')(decoder0)
 
-	model = models.Model(inputs=[inputs], outputs=[outputs])
+    model = models.Model(inputs=[inputs], outputs=[outputs])
 
-	model.compile(
+    model.compile(
             optimizer=optim, 
             loss = weighted_bce,
             #loss=losses.get(LOSS),
             metrics=[metrics.get(metric) for metric in mets])
 
-	return model
+    return model