Add doc for hw-frame-extract.

core/preprocess/hwvideoframe/README.md

 # hwvideoframe
+
 Hwvideoframe is a CV preprocessing library based on cuda. The project uses GPU for image preprocessing operations. It speeds up the processing speed while increasing the utilization rate of the GPU.
 
 ## Preprocess API
+
 Hwvideoframe provides a variety of data preprocessing methods for photo preprocess：
 
 - class Image2Gpubuffer
@@ -31,8 +33,8 @@ Hwvideoframe provides a variety of data preprocessing methods for photo preproce
 - class Normalize
 
   - `__init__(mean,std)`
-    - mean（list）：Length of list must be three.
-    - std（list）：Length of list must be three.
+    - mean（list）：Mean. Length of list must be three.
+    - std（list）：Variance. Length of list must be three.
   - `__call__(img)`
     - img（np.array）：Image data in (C,H,W) channels.
 
@@ -51,7 +53,9 @@ Hwvideoframe provides a variety of data preprocessing methods for photo preproce
     - img（numpy array）：Image data in (C,H,W) channels.
 
 ## Quick start
+
 [After compiling from code](https://github.com/PaddlePaddle/Serving/blob/develop/doc/COMPILE.md)，this project will be stored in reader。
 
 ## How to Test
+
 Test file：Serving/python/paddle_serving_app/reader/test_preprocess.py

core/preprocess/nvdec-extractframe/README.md

 # hw-frame-extract
+hw-frame-extract is a CV preprocessing library based on cuda. The project uses GPU for extracting frame from video. It speeds up the processing speed while increasing the utilization rate of the GPU.
+
+## Preprocess API
+Hwvideoframe provides a variety of cutting frame methods for different kind of video：
+
+- class HwExtractFrameJpeg
+
+  - `init_handler(gpu_index)`
+    - gpu_index (int)：Index of gpu.
+  - `extract_frame(file_path, n)`
+    - file_path (string): Video path.
+    - n (int): Cut n frames in one second.
+
+- class HwExtractFrameBGRARaw
+
+  - `init_handler(gpu_index)`
+    - gpu_index (int)：Index of gpu.
+  - `extract_frame(file_path, n)`
+    - file_path (string): Video path.
+    - n (int): Cut n frames in one second.
+
+- class HwFrameResult
+
+  - `height()`
+    - Return height of frame.
+  - `width()`
+    - Return width of frame.
+  - `len()`
+    - Return len of frame.
+  - `free_memory()`
+    - Free memory of frames.
+
+## Quick start
+[After compiling from code](https://github.com/PaddlePaddle/Serving/blob/develop/doc/COMPILE.md)，this project will be stored in reader。
+
+## How to Test
+Test file：Serving/python/paddle_serving_app/reader/test_preprocess.py

python/paddle_serving_app/reader/__init__.py

@@ -20,4 +20,4 @@ from .lac_reader import LACReader
 from .senta_reader import SentaReader
 from libgpupreprocess import Image2Gpubuffer, Gpubuffer2Image, RGB2BGR, BGR2RGB
 from libgpupreprocess import Div, Sub, Normalize, CenterCrop, Resize, ResizeByFactor
-from libhwextract import HwExtractFrameJpeg, HwExtractFrameBGRARaw
+from libhwextract import HwExtractFrameJpeg, HwExtractFrameBGRARaw, HwFrameResult