lisawen commited on
Commit
adcdf8a
1 Parent(s): f4ee313
Files changed (1) hide show
  1. soybean_dataset.py +22 -48
soybean_dataset.py CHANGED
@@ -117,69 +117,43 @@ class SoybeanDataset(datasets.GeneratorBasedBuilder):
117
  name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["valid"]}),
118
  ]
119
 
120
- def process_image(self,image_url):
121
- response = requests.get(image_url)
122
- response.raise_for_status() # This will raise an exception if there is a download error
123
-
124
- # Open the image from the downloaded bytes and return the PIL Image
125
- img = Image.open(BytesIO(response.content))
126
- return img
127
-
128
-
129
 
130
  def _generate_examples(self, filepath):
131
- #"""Yields examples as (key, example) tuples."""
132
  logging.info("generating examples from = %s", filepath)
133
 
134
  with open(filepath, encoding="utf-8") as f:
135
  data = csv.DictReader(f)
136
 
137
-
138
  for row in data:
139
- # Assuming the 'original_image' column has the full path to the image file
140
  unique_id = row['unique_id']
141
- original_image_path = row['original_image']
142
- segmentation_image_path = row['segmentation_image']
143
- sets = row['sets']
144
-
145
- original_image = self.process_image(original_image_path)
146
- segmentation_image = self.process_image(segmentation_image_path)
147
-
148
-
149
- # Here you need to replace 'initial_radius', 'final_radius', 'initial_angle', 'final_angle', 'target'
150
- # with actual columns from your CSV or additional processing you need to do
151
- yield row['unique_id'], {
 
 
 
 
 
 
152
  "unique_id": unique_id,
153
- "sets": sets,
154
  "original_image": original_image,
155
  "segmentation_image": segmentation_image,
156
  # ... add other features if necessary
157
  }
158
 
159
- # for row in data:
160
- # # Assuming the 'original_image' column has the full path to the image file
161
- # unique_id = row['unique_id']
162
- # original_image_path = row['original_image']
163
- # segmentation_image_path = row['segmentation_image']
164
- # sets = row['sets']
165
-
166
- # original_image_array = self.process_image(original_image_path)
167
- # segmentation_image_array = self.process_image(segmentation_image_path)
168
-
169
-
170
- # # Here you need to replace 'initial_radius', 'final_radius', 'initial_angle', 'final_angle', 'target'
171
- # # with actual columns from your CSV or additional processing you need to do
172
- # yield row['unique_id'], {
173
- # "unique_id": unique_id,
174
- # "sets": sets,
175
- # "original_image": original_image_array,
176
- # "segmentation_image": segmentation_image_array,
177
- # # ... add other features if necessary
178
- # }
179
-
180
-
181
-
182
-
183
 
184
 
185
 
 
117
  name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["valid"]}),
118
  ]
119
 
120
+ def process_image(self, image_path):
121
+ # Load the image from the local filesystem
122
+ img = Image.open(image_path)
123
+ return img
 
 
 
 
 
124
 
125
  def _generate_examples(self, filepath):
 
126
  logging.info("generating examples from = %s", filepath)
127
 
128
  with open(filepath, encoding="utf-8") as f:
129
  data = csv.DictReader(f)
130
 
 
131
  for row in data:
 
132
  unique_id = row['unique_id']
133
+ original_image_path = row['original_image'] # Adjust this path if necessary
134
+ segmentation_image_path = row['segmentation_image'] # Adjust this path if necessary
135
+
136
+ # Check if image exists locally before loading
137
+ if os.path.exists(original_image_path):
138
+ original_image = self.process_image(original_image_path)
139
+ else:
140
+ logging.error(f"Original image not found: {original_image_path}")
141
+ continue # or handle missing image appropriately
142
+
143
+ if os.path.exists(segmentation_image_path):
144
+ segmentation_image = self.process_image(segmentation_image_path)
145
+ else:
146
+ logging.error(f"Segmentation image not found: {segmentation_image_path}")
147
+ continue # or handle missing image appropriately
148
+
149
+ yield unique_id, {
150
  "unique_id": unique_id,
151
+ "sets": row['sets'],
152
  "original_image": original_image,
153
  "segmentation_image": segmentation_image,
154
  # ... add other features if necessary
155
  }
156
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
157
 
158
 
159