在帧中搜索关键词 Args: frame: 帧数据 keyword: 要搜索的关键词 enable_ocr: 是否启用OCR识别 Returns: bool: 是否找到关键词
(self, frame: Dict[str, Any], keyword: str, enable_ocr: bool = True)
| 387 | return self.process_text(combined_text) |
| 388 | |
| 389 | def match_keyword_in_frame(self, frame: Dict[str, Any], keyword: str, enable_ocr: bool = True) -> bool: |
| 390 | """ |
| 391 | 在帧中搜索关键词 |
| 392 | |
| 393 | Args: |
| 394 | frame: 帧数据 |
| 395 | keyword: 要搜索的关键词 |
| 396 | enable_ocr: 是否启用OCR识别 |
| 397 | |
| 398 | Returns: |
| 399 | bool: 是否找到关键词 |
| 400 | """ |
| 401 | if not keyword: |
| 402 | return False |
| 403 | |
| 404 | # 首先在现有文本字段中搜索 |
| 405 | text_fields = ['text', 'task_description', 'reasoning'] |
| 406 | for field in text_fields: |
| 407 | if field in frame and frame[field]: |
| 408 | processed = self.process_text(str(frame[field])) |
| 409 | if self.smart_text_contains(processed, keyword): |
| 410 | logger.debug(f"在字段 {field} 中找到关键词: {keyword}") |
| 411 | return True |
| 412 | |
| 413 | # 在XML文本中搜索 |
| 414 | if 'xml_text' in frame and frame['xml_text']: |
| 415 | xml_text = self.extract_xml_text(frame['xml_text']) |
| 416 | if xml_text: |
| 417 | processed = self.process_text(xml_text) |
| 418 | if self.smart_text_contains(processed, keyword): |
| 419 | logger.debug(f"在XML文本中找到关键词: {keyword}") |
| 420 | return True |
| 421 | |
| 422 | # 使用OCR在图像中搜索 |
| 423 | if enable_ocr and 'image' in frame and frame['image']: |
| 424 | if os.path.exists(frame['image']): |
| 425 | ocr_text, backup_text = self.extract_text_from_image(frame['image']) |
| 426 | |
| 427 | # 在主OCR结果中搜索 |
| 428 | if ocr_text: |
| 429 | processed = self.process_text(ocr_text) |
| 430 | if self.smart_text_contains(processed, keyword): |
| 431 | logger.debug(f"在OCR文本中找到关键词: {keyword}") |
| 432 | return True |
| 433 | |
| 434 | # 在备用OCR结果中搜索 |
| 435 | if backup_text and backup_text != ocr_text: |
| 436 | processed = self.process_text(backup_text) |
| 437 | if self.smart_text_contains(processed, keyword): |
| 438 | logger.debug(f"在备用OCR文本中找到关键词: {keyword}") |
| 439 | return True |
| 440 | |
| 441 | return False |
| 442 | |
| 443 | def get_text_similarity(self, text1: str, text2: str) -> float: |
| 444 | """计算两个文本的相似度""" |
no test coverage detected