File size: 3,324 Bytes
c701aec |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 |
import os
import json
import shutil
# 📂 **路径配置**
ANNOTATION_DIR = r"C:\Users\yang2\Desktop\YVT\holisticOcrYvt"
VIDEO_DIR = r"C:\Users\yang2\Desktop\YVT\YVT\videos\test"
VIDEO_DST_DIR = r"C:\Users\yang2\Desktop\BenchMark\HolisticOCR\Videos1"
ANNOTATION_OUTPUT_PATH = r"C:\Users\yang2\Desktop\BenchMark\HolisticOCR\holisticOCRTest1.json"
# 确保目标路径存在
os.makedirs(VIDEO_DST_DIR, exist_ok=True)
def process_files(annotation_dir, video_dir, video_dst_dir, output_json):
"""
处理 YVT 数据集，将视频和标注文件统一重命名并合并 JSON 标注。
"""
combined_annotations = []
question_id = 551
# 如果文件已存在，先读取已有标注，防止覆盖
if os.path.exists(output_json):
with open(output_json, 'r', encoding='utf-8') as f:
combined_annotations = json.load(f)
existing_ids = {qa['question_id'] for qa in combined_annotations}
question_id = max(existing_ids) + 1 if existing_ids else 551
for file in os.listdir(annotation_dir):
if file.endswith(".json"):
annotation_path = os.path.join(annotation_dir, file)
# 去掉.mp4_holisticOcr后缀，匹配视频名
video_base_name = file.replace(".mp4_holisticOcr.json", "")
video_src = os.path.join(video_dir, f"{video_base_name}.mp4")
video_dst = os.path.join(video_dst_dir, f"video_{question_id}.mp4")
if os.path.exists(video_src):
# 1️⃣ **复制并重命名视频**
shutil.copy(video_src, video_dst)
print(f"✅ Video copied: {video_src} -> {video_dst}")
# 2️⃣ **读取 JSON 标注**
with open(annotation_path, 'r', encoding='utf-8') as f:
annotation_data = json.load(f)
text_content = annotation_data.get('answer', '')
text_answer = " ".join(text_content.split()) # 确保文本格式统一
# 3️⃣ **生成新的 JSON 标注**
annotation_entry = {
"video_id": f"video_{question_id}",
"question": "请给出视频中所有文本内容。",
"answer": text_answer,
"type": "holisticOCR",
"source": video_base_name,
"question_id": question_id
}
combined_annotations.append(annotation_entry)
print(f"✅ Annotation added: video_{question_id}, source: {video_base_name}")
question_id += 1
else:
print(f"⚠️ Video not found: {video_src}")
# 4️⃣ **保存合并后的 JSON 文件**
with open(output_json, 'w', encoding='utf-8') as out_f:
json.dump(combined_annotations, out_f, ensure_ascii=False, indent=4)
print(f"🎯 Annotations updated and saved to {output_json}")
if __name__ == "__main__":
try:
process_files(ANNOTATION_DIR, VIDEO_DIR, VIDEO_DST_DIR, ANNOTATION_OUTPUT_PATH)
print("🚀 YVT Holistic OCR Benchmark 数据集重命名与整合完成！")
except Exception as e:
print(f"❌ 发生错误: {e}")
|