加点字，斜体字,划横线的字,删除线的字识别格式丢失

[yukaige]

加点字，斜体字,划横线的字,删除线的字识别格式丢失

[pratikrath126]

PaddleOCR focuses on text detection and recognition it does not preserve typographic formatting (bold, italic, underline, strikethrough). Those styles are not part of the recognized text output. You would need additional processing like analyzing stroke width, using document layout models (e.g., PP-Structure for key-value extraction) or custom post-processing to infer styling from the image. At the moment, PaddleOCR does not support style detection out-of-the-box.