diff --git a/doc/table/tableocr_pipeline.jpg b/doc/table/tableocr_pipeline.jpg
new file mode 100644
index 0000000000..bd467b1bd3
Binary files /dev/null and b/doc/table/tableocr_pipeline.jpg differ
diff --git a/doc/table/tableocr_pipeline.png b/doc/table/tableocr_pipeline.png
deleted file mode 100644
index 731b84da9b..0000000000
Binary files a/doc/table/tableocr_pipeline.png and /dev/null differ
diff --git a/doc/table/tableocr_pipeline_en.jpg b/doc/table/tableocr_pipeline_en.jpg
new file mode 100644
index 0000000000..654366878e
Binary files /dev/null and b/doc/table/tableocr_pipeline_en.jpg differ
diff --git a/ppstructure/table/README.md b/ppstructure/table/README.md
index b0692769a0..afcbe1696b 100644
--- a/ppstructure/table/README.md
+++ b/ppstructure/table/README.md
@@ -8,7 +8,7 @@ The ocr of the table mainly contains three models
 
 The table ocr flow chart is as follows
 
-![tableocr_pipeline](../../doc/table/tableocr_pipeline.png)
+![tableocr_pipeline](../../doc/table/tableocr_pipeline_en.jpg)
 
 1. The coordinates of single-line text is detected by DB model, and then sends it to the recognition model to get the recognition result.
 2. The table structure and cell coordinates is predicted by RARE model.
diff --git a/ppstructure/table/README_ch.md b/ppstructure/table/README_ch.md
index 1f2b8b1dee..4b912f3eb8 100644
--- a/ppstructure/table/README_ch.md
+++ b/ppstructure/table/README_ch.md
@@ -8,7 +8,7 @@
 
 具体流程图如下
 
-![tableocr_pipeline](../../doc/table/tableocr_pipeline.png)
+![tableocr_pipeline](../../doc/table/tableocr_pipeline.jpg)
 
 1. 图片由单行文字检测检测模型到单行文字的坐标，然后送入识别模型拿到识别结果。
 2. 图片由表格结构和cell坐标预测模型拿到表格的结构信息和单元格的坐标信息。
@@ -17,7 +17,6 @@
 
 ## 2. 使用
 
-
 ### 2.1 训练
 在这一章节中，我们仅介绍表格结构模型的训练，[文字检测](../../doc/doc_ch/detection.md)和[文字识别](../../doc/doc_ch/recognition.md)的模型训练请参考对应的文档。