diff --git a/README.md b/README.md
index ed01dafdb9b8317a163172a49d68fc40559d83a2..c6ecd489d3de999998573c77e2de160ab0816335 100644
--- a/README.md
+++ b/README.md
@@ -13,9 +13,14 @@ PaddleOCR旨在打造一套丰富、领先、且实用的OCR工具库，助力
 - 支持模型部署
 
 ![](./doc/imgs_draw/11.jpg)
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/11.jpg"/>
+
+注：更多效果展示请见文末。
 
 ## **快速运行**
 
+快速运行前请先参考[快速安装](./doc/installation.md)配置PaddleOCR工作环境。
+
 下载inference模型
 ```
 # 创建inference模型保存目录
@@ -90,7 +95,16 @@ PaddleOCR文本识别算法的训练与使用请参考[文档](./doc/recognition
 PaddleOCR即将开源百度自研端对端OCR模型[End2End-PSL](https://arxiv.org/abs/1909.07808)，敬请关注。
 - [ ]  End2End-PSL (百度自研, comming soon)
 
+## 效果展示
 
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/1.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/4.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/6.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/7.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/9.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/12.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/16.jpg"/>
+<img width="500" height="300" src="https://github.com/PaddlePaddle/PaddleOCR/doc/imgs_draw/22.jpg"/>
 
 # 参考文献
 ```
diff --git a/doc/imgs/18.jpg b/doc/imgs/18.jpg
deleted file mode 100644
index f1c8b8d099bdfbc7290e430dc2b97c842b9d65bf..0000000000000000000000000000000000000000
Binary files a/doc/imgs/18.jpg and /dev/null differ
diff --git a/doc/imgs_draw/18.jpg b/doc/imgs_draw/18.jpg
deleted file mode 100644
index 729e2dc5797babf7cdf38fc5e3ea049874508b0c..0000000000000000000000000000000000000000
Binary files a/doc/imgs_draw/18.jpg and /dev/null differ