Скрипт для валидации датасета.

This commit is contained in:
2025-07-07 18:39:15 +03:00
parent 76b0919e91
commit e2047bbf5a
2 changed files with 36 additions and 0 deletions

1
.gitignore vendored
View File

@@ -33,3 +33,4 @@ paddleocr.egg-info/
/deploy/android_demo/app/cache/
test_tipc/web/models/
test_tipc/web/node_modules/
paddle/

View File

@@ -0,0 +1,35 @@
import os
import cv2
def rename_files(directory_path: str):
extensions = set()
for file in os.listdir(directory_path):
filename, extension = file.split(".")
extensions.add(extension)
old_filename = os.path.join(directory_path, file)
new_filename = os.path.join(directory_path, f"{filename}.jpg")
if old_filename != new_filename:
print(f"{old_filename} -> {new_filename}")
os.rename(old_filename, new_filename)
def check_images(directory_path: str):
for file in os.listdir(directory_path):
try:
img = cv2.imread(os.path.join(directory_path, file))
if img is None:
print(file)
except Exception:
print(file)
def check_labels(dir_path: str) -> int:
for filename in os.listdir(dir_path):
name, extension = filename.split(".")
if len(name) != 17:
print(filename)
# rename_files("train_data/images/")
# check_images("train_data/images/")
check_labels("train_data/images/")