keyerror when trying to load data - parsing

Question

I was trying to parse my dataset with the images and annotations with the code below. I am using anaconda with Spyder IDE in a windows machine:
DATA_DIR = 'input'
# Directory to save logs and trained model
ROOT_DIR = 'working'
train_dicom_dir = os.path.join(DATA_DIR, 'stage_1_train_images')
test_dicom_dir = os.path.join(DATA_DIR, 'stage_1_test_images')
print(train_dicom_dir)
print(test_dicom_dir)
def get_dicom_fps(dicom_dir):
dicom_fps = glob.glob(dicom_dir+'/'+'*.jpg')
return list(set(dicom_fps))
def parse_dataset(dicom_dir, anns):
image_fps = get_dicom_fps(dicom_dir)
image_annotations = {fp: [] for fp in image_fps}
for index, row in anns.iterrows():
fp = os.path.join(dicom_dir, row['patientId']+'.jpg')
image_annotations[fp].append(row)
return image_fps, image_annotations
image_fps, image_annotations = parse_dataset(train_dicom_dir, anns=anns)
On running the code, i get the following error:
Traceback (most recent call last):
File "<ipython-input-21-fe2564bb2360>", line 1, in <module>
image_fps, image_annotations = parse_dataset(train_dicom_dir, anns=anns)
File "<ipython-input-12-a5d26437db38>", line 6, in parse_dataset
image_annotations[fp].append(row)
KeyError: 'input\\stage_1_train_images\\1.2.276.0.7230010.3.1.4.8323329.10001.1517874346.163716.jpg'
My current working directory is'C:\Users\rajaramans2\codes\input'

Answer 1

It seems like you are trying to process images in .dcm (DICOM format), not .jpg, so your code should work if you replace “.jpg” for “.dcm”, I hope this helps.