Skip to content

Commit

Permalink
update
Browse files Browse the repository at this point in the history
  • Loading branch information
@CUQS
CUQS committed Dec 23, 2019
1 parent 0a2c449 commit 701dedc
Show file tree
Hide file tree
Showing 12 changed files with 182 additions and 113 deletions.
126 changes: 80 additions & 46 deletions README.md
View file
Original file line number Diff line number Diff line change
@@ -1,38 +1,74 @@
EN|[CN](README_cn.md)

# Road Segmentation
# ObjDetection<a name="EN-US_TOPIC_0167609448"></a>

This application runs on the Atlas 200 DK and implements the road segmentation by using fcn network.
The object detection application runs on the Atlas 200 DK and implements the inference function by using Faster R-CNN object detection network.

<p align='center'>
<img src='img/img3.jpg' height=300>
</p>

<p align='center'>
<img src='to_out/test.png'>
<img src='img/pictest.png' width=300>
<img stc='img/realTime.png' width=300>
</p>

<p align='center'>
<img src='img/img4.png'>
</p>

## Prerequisites
## Prerequisites<a name="en-us_topic_0182554604_section137245294533"></a>

Before using an open source application, ensure that:

- Mind Studio has been installed.
- The Atlas 200 DK developer board has been connected to Mind Studio, the cross compiler has been installed, the SD card has been prepared, and basic information has been configured.

## Software Preparation
## Software Preparation<a name="en-us_topic_0182554604_section181111827718"></a>

Before running the application, obtain the source code package and configure the environment as follows.

1. Obtain the source code package.
1. Obtain the source code package.

Download all the code in the objDetection repository at [https://github.com/Ascend/objDetection](https://github.com/Ascend/objDetection) to any directory on Ubuntu Server where Mind Studio is located as the Mind Studio installation user, for example, _/home/ascend/objdDetection_.

2. <a name="en-us_topic_0182554604_li2074865610364"></a>Obtain the source network model required by the application.

Obtain the source network model and its weight file used in the application by referring to [Table 1](#en-us_topic_0182554604_table19942111763710), and save them to any directory on the Ubuntu server where Mind Studio is located (for example, **$HOME/ascend/models/faster\_rcnn**).

**Table 1** Models used in the detection network application

<a name="en-us_topic_0182554604_table19942111763710"></a>
<table><thead align="left"><tr id="en-us_topic_0182554604_row611318123710"><th class="cellrowborder" valign="top" width="11.959999999999999%" id="mcps1.2.4.1.1"><p id="en-us_topic_0182554604_p81141820376"><a name="en-us_topic_0182554604_p81141820376"></a><a name="en-us_topic_0182554604_p81141820376"></a>Model Name</p>
</th>
<th class="cellrowborder" valign="top" width="8.07%" id="mcps1.2.4.1.2"><p id="en-us_topic_0182554604_p13181823711"><a name="en-us_topic_0182554604_p13181823711"></a><a name="en-us_topic_0182554604_p13181823711"></a>Model Description</p>
</th>
<th class="cellrowborder" valign="top" width="79.97%" id="mcps1.2.4.1.3"><p id="en-us_topic_0182554604_p1717182378"><a name="en-us_topic_0182554604_p1717182378"></a><a name="en-us_topic_0182554604_p1717182378"></a>Model Download Path</p>
</th>
</tr>
</thead>
<tbody><tr id="en-us_topic_0182554604_row1119187377"><td class="cellrowborder" valign="top" width="11.959999999999999%" headers="mcps1.2.4.1.1 "><p id="en-us_topic_0182554604_p7118189378"><a name="en-us_topic_0182554604_p7118189378"></a><a name="en-us_topic_0182554604_p7118189378"></a>faster_rcnn</p>
</td>
<td class="cellrowborder" valign="top" width="8.07%" headers="mcps1.2.4.1.2 "><p id="en-us_topic_0182554604_p151818183718"><a name="en-us_topic_0182554604_p151818183718"></a><a name="en-us_topic_0182554604_p151818183718"></a>Network model for object detection.</p>
<p id="en-us_topic_0182554604_p8394945195815"><a name="en-us_topic_0182554604_p8394945195815"></a><a name="en-us_topic_0182554604_p8394945195815"></a>It is a Faster R-CNN model based on Caffe.</p>
</td>
<td class="cellrowborder" valign="top" width="79.97%" headers="mcps1.2.4.1.3 "><p id="en-us_topic_0182554604_p611318163718"><a name="en-us_topic_0182554604_p611318163718"></a><a name="en-us_topic_0182554604_p611318163718"></a>Download the source network model file and its weight file by referring to<strong id="en-us_topic_0182554604_b17606155113121"><a name="en-us_topic_0182554604_b17606155113121"></a><a name="en-us_topic_0182554604_b17606155113121"></a> README.md</strong> in <a href="https://github.com/Ascend/models/tree/master/computer_vision/object_detect/faster_rcnn" target="_blank" rel="noopener noreferrer">https://github.com/Ascend/models/tree/master/computer_vision/object_detect/faster_rcnn</a>.</p>
</td>
</tr>
</tbody>
</table>

3. Convert the source network model to a Da Vinci model.
1. Choose **Tool \> Convert Model** from the main menu of Mind Studio. The **Convert Model** page is displayed.
2. On the **Convert Model** page, set **Model File** and **Weight File** to the model file and weight file downloaded in [2](#en-us_topic_0182554604_li2074865610364), respectively.
- Set **Model Name** to the model name in [Table 1](#en-us_topic_0182554604_table19942111763710): **faster\_rcnn**.
- Retain default values for other parameters.

Download all the code in the road-segmentation repository at [https://github.com/Ascend/road-segmentation](https://github.com/Ascend/road-segmentation) to any directory on Ubuntu Server where Mind Studio is located as the Mind Studio installation user, for example, _/home/ascend/road-segmentation_.
**Figure 1** faster\_rcnn model conversion configuration<a name="en-us_topic_0182554604_fig79252510407"></a>
![](img/faster_rcnn-model-conversion-configuration.jpg "faster_rcnn-model-conversion-configuration")


2. Log in to Ubuntu Server where Mind Studio is located as the Mind Studio installation user and set the environment variable **DDK\_HOME**.
3. Click **OK** to start model conversion.

After successful conversion, a .om Da Vinci model file is generated in the **$HOME/tools/che/model-zoo/my-model/fast\_rcnn** directory.

4. Log in to Ubuntu Server where Mind Studio is located as the Mind Studio installation user and set the environment variable **DDK\_HOME**.

**vim \~/.bashrc**

Expand All @@ -43,48 +79,49 @@ Before running the application, obtain the source code package and configure the
**export LD\_LIBRARY\_PATH=$DDK\_HOME/uihost/lib**

>**NOTE:**
>
>- **XXX** indicates the Mind Studio installation user, and **/home/XXX/tools** indicates the default installation path of the DDK.
>- If the environment variables have been added, skip this step.
Enter **:wq!** to save and exit.

Run the following command for the environment variable to take effect:

>- If the environment variables have been added, skip this step.
Enter **:wq!** to save and exit.
Run the following command for the environment variable to take effect:
**source \~/.bashrc**


## Deployment
## Deployment<a name="en-us_topic_0182554604_section3723145213347"></a>

1. Access the root directory where the application code is located as the Mind Studio installation user, for example, **_/home/ascend/road-segmentation_**.
1. Access the root directory where the cvnetworkverify application code is located as the Mind Studio installation user, for example, **_/home/ascend/objDetection_**.
2. Run the deployment script to prepare the project environment, including compiling and deploying the ascenddk public library and application.

bash deploy.sh _host\_ip_ _model\_mode_

- _host\_ip_: For the Atlas 200 DK developer board, this parameter indicates the IP address of the developer board.
- _model\_mode_ indicates the deployment mode of the model file. The default setting is **internet**.
- **local**: If the Ubuntu system where Mind Studio is located is not connected to the network, use the local mode. In this case, download the dependent common code library ezdvpp to the **sample-objectdetection/script** directory by referring to the [Downloading Dependent Code Library](#en-us_topic_0182554604_section92241245122511).
- **local**: If the Ubuntu system where Mind Studio is located is not connected to the network, use the local mode. In this case, download the dependent common code library ezdvpp to the **objDetection/script** directory by referring to the [Downloading Dependent Code Library](#en-us_topic_0182554604_section92241245122511).
- **internet**: Indicates the online deployment mode. If the Ubuntu system where Mind Studio is located is connected to the network, use the Internet mode. In this case, download the dependent code library ezdvpp online.


Example command:

**bash deploy.sh 192.168.1.2**

3. Upload the generated Da Vinci offline model **kittisegRealTime.om** to the directory of the **HwHiAiUser** user on the host.
3. Upload the generated Da Vinci offline model **faster_rcnn.om** to the directory of the **HwHiAiUser** user on the host.

```bash
scp kittisegRealTime.om HwHiAiUser@host_ip:/home/HwHiAiUser/HIAI_PROJECTS/ascend_workspace/segmentation/out/kittisegRealTime.om
```
```bash
scp faster_rcnn.om HwHiAiUser@host_ip:/home/HwHiAiUser/HIAI_PROJECTS/ascend_workspace/objDetection/out/faster_rcnn.om
```

For the Atlas 200 DK, the default value of _**host\_ip**_ is **192.168.1.2** \(USB connection mode\) or **192.168.0.2** \(NIC connection mode\).
For the Atlas 200 DK, the default value of _**host\_ip**_ is **192.168.1.2** \(USB connection mode\) or **192.168.0.2** \(NIC connection mode\).


## Running
## Running<a name="en-us_topic_0182554604_section87121843104920"></a>

1. run server on the host

```bash
python3 run_server.py
python3 run_server.py
```

2. Log in to the Host as the **HwHiAiUser** user in SSH mode on Ubuntu Server where Mind Studio is located.
Expand All @@ -95,36 +132,32 @@ Before running the application, obtain the source code package and configure the

For the Atlas 200 DK, the default value of _**host\_ip**_ is **192.168.1.2** \(USB connection mode\) or **192.168.0.2** \(NIC connection mode\).

3. Go to the path of the executable file of road segmentation application.
3. Go to the path of the executable file of road segmentation application.

**cd \~/HIAI\_PROJECTS/ascend\_workspace/segmentation/out**
**cd \~/HIAI\_PROJECTS/ascend\_workspace/objDetection/out**

4. Run the application.
4. Run the application.

Camera Test

![image2](img/realTime.png)

5. Picture Test

Camera Test

```bash
./ascend_segmentation
```
![image2](img/img2.png)

Picture Test

```bash
./ascend_segmentation 1
```
![image1](img/img1.png)

- Width of the input image: 623px
- Height of the input image: 188px

![pic](img/pictest.png)


## Downloading Dependent Code Library<a name="en-us_topic_0182554604_section92241245122511"></a>

Download the dependent software libraries to the road-segmentation/script** directory.
Download the dependent software libraries to the **objDetection/script** directory.

**Table 2** Download the dependent software library

<a name="en-us_topic_0182554604_table6701646132617"></a>
<table><thead align="left"><tr id="en-us_topic_0182554604_row177421045163614"><th class="cellrowborder" valign="top" width="33.33333333333333%" id="mcps1.2.4.1.1"><p id="en-us_topic_0182554604_p574264511368"><a name="en-us_topic_0182554604_p574264511368"></a><a name="en-us_topic_0182554604_p574264511368"></a>Module Name</p>
</th>
<th class="cellrowborder" valign="top" width="33.33333333333333%" id="mcps1.2.4.1.2"><p id="en-us_topic_0182554604_p1474224573615"><a name="en-us_topic_0182554604_p1474224573615"></a><a name="en-us_topic_0182554604_p1474224573615"></a>Module Description</p>
Expand All @@ -143,3 +176,4 @@ Download the dependent software libraries to the road-segmentation/script** di
</tr>
</tbody>
</table>

Loading

0 comments on commit 701dedc

Please sign in to comment.