Difference between revisions of "MV Camera on Orange Pi's RK35XX Boards"

From wiki_veye
Jump to navigation Jump to search
 
(9 intermediate revisions by 2 users not shown)
Line 1: Line 1:
 
[https://wiki.veye.cc/index.php/MV_Camera_on_Orange_Pi%27s_RK35XX_Boards/zh 查看中文]
 
[https://wiki.veye.cc/index.php/MV_Camera_on_Orange_Pi%27s_RK35XX_Boards/zh 查看中文]
  
−
'''<big>怎样在Orange Pi的RK35XX板子上使用MV系列和RAW系列相机</big>'''
+
===Overview===
−
===概述 ===
+
The MV series and RAW series cameras are cameras designed for AI applications in the industrial field. They use the MIPI CSI-2 interface and are particularly suitable for use with embedded computing platforms. They have rich data formats and triggering features, extremely low latency, high bandwidth, and reliable stability.
−
MV系列和RAW系列相机是为工业领域的AI应用推出的相机。其使用MIPI CSI-2接口，尤其适合嵌入式运算平台使用。它具备丰富的数据格式和触发特性，极低的延时，极高的带宽和可靠的稳定性。
 
  
−
本文以Orange Pi的CM4和CM5主板为例，介绍怎样将MV和RAW系列摄像机接入RK3566/3K3568系统。
+
This article takes OrangePi CM4 and OrangePi CM5 and CM5 tablet board as an example to introduce how to connect MV and RAW series cameras to the RK3566/3K3568 and RK3588S/RK3588 system.
  
−
我们提供了Linux操作系统下的驱动。
+
We provide drivers for the Linux operating system (using Ubuntu as an example).
−
====支持的模组====
+
====Camera Module List====
 
{| class="wikitable"
 
{| class="wikitable"
−
!系列
+
!Series
−
!型号
+
!Model
−
!状态
+
!Status
 
|-
 
|-
−
|MV系列
+
|MV series
 
|MV-MIPI-IMX178M
 
|MV-MIPI-IMX178M
−
|完成
+
|Done
 
|-
 
|-
−
|MV系列
+
|MV series
 
|MV-MIPI-SC130M
 
|MV-MIPI-SC130M
−
|完成
+
|Done
 
|-
 
|-
−
|MV系列
+
|MV series
 +
|MV-MIPI-IMX296M
 +
|Done
 +
|-
 +
|MV series
 
|MV-MIPI-IMX287M
 
|MV-MIPI-IMX287M
−
|完成
+
|Done
 
|-
 
|-
−
|MV系列
+
|MV series
−
|MV-MIPI-IMX296M
 
−
|完成
 
−
|-
 
−
|MV系列
 
 
|MV-MIPI-IMX265M
 
|MV-MIPI-IMX265M
−
|完成
+
|Done
 
|-
 
|-
−
|MV系列
+
|MV series
 
|MV-MIPI-IMX264M
 
|MV-MIPI-IMX264M
−
|完成
+
|Done
 
|-
 
|-
−
|RAW系列
+
|MV series
 +
|MV-MIPI-GMAX4002M
 +
|Done
 +
|-
 +
|RAW series
 
|RAW-MIPI-SC132M
 
|RAW-MIPI-SC132M
−
|完成
+
|Done
 
|-
 
|-
−
|RAW系列
+
|RAW series
 
|RAW-MIPI-IMX462M
 
|RAW-MIPI-IMX462M
−
|完成
+
|Done
 
|-
 
|-
−
|RAW系列
+
|RAW series
 
|RAW-MIPI-AR0234M
 
|RAW-MIPI-AR0234M
−
|完成
+
|Done
 
|-
 
|-
−
|RAW系列
+
|RAW series
 
|RAW-MIPI-SC535M
 
|RAW-MIPI-SC535M
−
|完成
+
|Done
 
|}
 
|}
−
===硬件准备及安装===
 
−
我们使用Orange Pi的CM4和CM5的官方底板，这两个底板提供了兼容树莓派的15Pin端子。对于RAW系列相机，无需使用转接板即可将我们的相机安装到其主板上。对于MV系列相机，则需要使用[[ADP-MV1 Adapter Board Data Sheet/zh|ADP-MV1]]转接板进行连接。
 
  
−
==== 相机与OrangePi CM4的连接 ====
+
===Hardware Setup===
−
ADP-MV1和Zero 3W之间使用15P的FFC异面线连接，注意接触面方向。
+
We use the official baseboards of the Orange Pi CM4 and CM5, which feature a 15-pin header compatible with Raspberry Pi. For the RAW series cameras, our cameras can be directly mounted onto the baseboard without the need for an adapter board. For the MV series cameras, the ADP-MV1 adapter board is required for connection.
−
[[File:OrangePi CM4 to MV cam.jpg|center|thumb|600x600px|OrangePi CM4 to MV cam]]
+
 
−
[[File:OrangePi CM4 to RAW cam.jpg|center|thumb|600x600px|OrangePi CM4 to RAW cam]]
+
====Camera Connection to Orange Pi CM4====
 +
The ADP-MV1 is connected to the OrangePi CM4 via a 15-pin FFC cable with opposite surface alignment; please pay attention to the orientation of the contact surfaces.
 +
 
 +
Note that only the CAM1 shown in the image below supports MV and RAW cameras.[[File:OrangePi CM4 to MV cam.jpg|center|thumb|600x600px|OrangePi CM4 to MV cam|link=http://wiki.veye.cc/index.php/File:OrangePi_CM4_to_MV_cam.jpg]][[File:OrangePi CM4 to RAW cam.jpg|center|thumb|600x600px|OrangePi CM4 to RAW cam|link=http://wiki.veye.cc/index.php/File:OrangePi_CM4_to_RAW_cam.jpg]]<br />
 +
====Camera Connection to Orange Pi CM5====
 +
The OrangePi CM5 supports up to four cameras. The following diagram shows the hardware connection method for simultaneously connecting multiple cameras.[[File:OrangePi CM5 to all cam overview.jpg|center|thumb|600x600px|OrangePi CM5 to all cam overview|link=http://wiki.veye.cc/index.php/File:OrangePi_CM5_to_all_cam_overview.jpg]][[File:OrangePi CM5 to all cam backview.jpg|center|thumb|600x600px|OrangePi CM5 to all cam backview|link=http://wiki.veye.cc/index.php/File:OrangePi_CM5_to_all_cam_backview.jpg]]<br />
 +
 
 +
====Camera Connection to Orange Pi CM5 tablet====
 +
Due to differences in connector formats, we have successfully interfaced modules for both CAM2 and CAM3 on the OrangePi CM5 tablet. The diagram below illustrates the connection methods for the RAW series and MV series.
 
<br />
 
<br />
 +
[[File:OrangePi CM5 tablet to MV cam.jpg|link=http://wiki.veye.cc/index.php/File:OrangePi%20CM5%20tablet%20to%20MV%20cam.jpg|center|thumb|600x600px|OrangePi CM5 tablet to RAW cam]]
 +
[[File:OrangePi CM5 tablet to MV cam .jpg|link=http://wiki.veye.cc/index.php/File:OrangePi%20CM5%20tablet%20to%20MV%20cam%20.jpg|center|thumb|600x600px|OrangePi CM5 tablet to MV cam]]
  
−
==== 相机与OrangePi CM5的连接 ====
+
===Introduction to github repositories===
−
OrangePi CM5最多支持4个VEYE摄像头，下图是同时连接多种摄像头的硬件连接方法展示。[[File:OrangePi CM5 to all cam overview.jpg|center|thumb|600x600px|OrangePi CM5 to all cam overview|link=http://wiki.veye.cc/index.php/File:OrangePi_CM5_to_all_cam_overview.jpg]][[File:OrangePi CM5 to all cam backview.jpg|center|thumb|600x600px|OrangePi CM5 to all cam backview|link=http://wiki.veye.cc/index.php/File:OrangePi_CM5_to_all_cam_backview.jpg]]
 
−
 
 
−
===github库介绍===
 
 
https://github.com/veyeimaging/rk35xx_veye_bsp
 
https://github.com/veyeimaging/rk35xx_veye_bsp
  
 
https://github.com/veyeimaging/rk35xx_orangepi
 
https://github.com/veyeimaging/rk35xx_orangepi
  
−
包括以下主要内容：
+
includes：
  
−
*driver和dts源码
+
*driver source code
−
*i2c通信工具集
+
*i2c toolkits
−
*应用层demo
+
*application demo
  
−
此外，在[https://github.com/veyeimaging/rk35xx_orangepi/releases releases]中，提供了编译好的烧写镜像和linux内核的deb安装包。
+
In addition, a compiled linux kernel installation package  is provided in the [https://github.com/veyeimaging/rk35xx_orangepi/releases releases].
−
===更新主板的Ubuntu系统===
+
===Upgrade the Ubuntu system===
−
我们提供了发布系统的烧写镜像，同时提供了linux内核的deb安装包。
+
We provide a flashing image for the release system, as well as a deb package for the Linux kernel.
  
−
参考OrangePi CM4[http://www.orangepi.cn/orangepiwiki/index.php/Orange_Pi_CM4 用户手册]，或者OrangePi CM5的[http://www.orangepi.cn/html/hardWare/computerAndMicrocontrollers/service-and-support/Orange-Pi-CM5.html 用户手册]烧写系统。当然，也可以使用通用的dpkg命令，安装deb包。
+
Refer to the [http://www.orangepi.org/orangepiwiki/index.php/Orange_Pi_CM4 OrangePi CM4 user manual] or the [http://www.orangepi.org/html/hardWare/computerAndMicrocontrollers/service-and-support/Orange-Pi-CM5.html OrangePi CM5 user manual] for instructions on flashing the system. Alternatively, you can use the general <code>dpkg</code> command to install the deb package.
−
===系统状态检测===
+
===Check system status===
−
完成系统安装后，系统上电。在主板端执行以下命令来检测是否正确连接相机。
+
Run the following command to confirm whether the camera is probed.
  
 
<code>sudo dmesg | grep mvcam</code>
 
<code>sudo dmesg | grep mvcam</code>
 
====CM4====
 
====CM4====
−
CM4只支持通过cam1接口，接入相机。以RAW-MIPI-SC132M为例，dmesg信息中包含如下内容
+
The CM4 supports camera connection only through the CAM1 interface. Taking the RAW-MIPI-SC132M as an example, the <code>dmesg</code> output contains the following information:
 +
 
 +
The output message appears as shown below：
  
 
<code>mvcam 1-003b: camera is: RAW-MIPI-SC132M</code>
 
<code>mvcam 1-003b: camera is: RAW-MIPI-SC132M</code>
Line 93: Line 103:
 
<code>mvcam 1-003b: firmware version: 0x1040000</code>
 
<code>mvcam 1-003b: firmware version: 0x1040000</code>
  
−
执行如下命令，检查video设备节点
+
*Run the following command to check the presence of video node.
  
 
<code>ls /dev/video0</code>
 
<code>ls /dev/video0</code>
  
−
可以看到
+
The output message appears as shown below.
  
 
<code>video0</code>
 
<code>video0</code>
  
−
正确识别摄像头之后，相机被识别为/dev/video0。
 
 
====CM5====
 
====CM5====
−
CM5 支持接入4个相机，以RAW-MIPI-SC132M为例，dmesg信息中包含如下内容dmesg信息中包含如下内容:
+
The CM5 supports the connection of up to four cameras. Taking the RAW-MIPI-SC132M as an example, the <code>dmesg</code> output contains the following information:
  
 
<code>mvcam 3-003b: camera is: RAW-MIPI-SC132M</code>
 
<code>mvcam 3-003b: camera is: RAW-MIPI-SC132M</code>
Line 120: Line 129:
  
 
<code>mvcam 6-003b: firmware version: 0x1040000</code>
 
<code>mvcam 6-003b: firmware version: 0x1040000</code>
−
====使用media-ctl查看拓扑结构====
+
====CM5 tablet====
−
下面以CM5的CAM1为例进行解释。
+
The CM5 supports the connection of up to two cameras. Taking the RAW-MIPI-SC132M as an example, the <code>dmesg</code> output contains the following information:
  
−
使用media-ctl指令，可以清晰的展现出当前的拓扑结构。
+
<code>mvcam 6-003b: camera is: RAW-MIPI-SC132M</code>
  
−
<code>media-ctl -p -d /dev/media2</code>
+
<code>mvcam 6-003b: firmware version: 0x1040000</code>
−
=====链接关系=====
 
−
mv camera->rockchip-csi2-dphy1->rockchip-mipi-csi2->stream_cif_mipi_id0 - - ->DDR(/dev/video22)
 
−
 
 
−
应用程序可以通过/dev/video22节点获取图像。
 
−
 
 
−
=====mv camera entity信息=====
 
−
以RAW-MIPI-SC132M为例：
 
−
 
 
−
<code>- entity 63: m00_b_mvcam 4-003b (1 pad, 1 link)</code>
 
−
 
 
−
<code>             type V4L2 subdev subtype Sensor flags 0</code>
 
−
 
 
−
<code>             device node name /dev/v4l-subdev8</code>
 
−
 
 
−
<code>        pad0: Source</code>
 
−
 
 
−
<code>                [fmt:Y8_1X8/1080x1280@100/12000 field:none]</code>
 
−
 
 
−
<code>                -> "rockchip-csi2-dphy1":0 [ENABLED]</code>
 
−
 
 
−
可以看到：
 
−
 
 
−
*该Entity完整的名称是： <code>m00_b_mvcam 4-003b</code>。
 
−
*它是一个 V4L2 subdev (Sub-Device) Sensor。
 
−
*它对应的节点是 <code>/dev/v4l-subdev8</code> ，应用程序（如<code>v4l2-ctl</code>）可以打开它，并进行配置。
 
−
*它的输出格式是 [<code>fmt:Y8_1X8/1080x1280@100/12000 field:none</code>] ，其中<code>Y8_1X8</code>是一种mbus-code的简写，下一小节会列出支持的mbus-code。
 
−
*当前分辨率是<code>1080x1280。</code>
 
−
*当前帧间隔是<code>100/12000</code>,即帧率是100。
 
−
 
 
−
摄像机输出的数据格式，可以通过media-ctl指令修改。
 
−
 
 
−
各个信息对应关系如下表：
 
−
 
 
−
- CM4
 
−
{| class="wikitable"
 
−
!CAM num
 
−
!I2C
 
−
!media node
 
−
!media entity name
 
−
!video node
 
−
!subdev node
 
−
|-
 
−
|1
 
−
|1
 
−
|/dev/media0
 
−
|m00_b_mvcam 1-003b
 
−
|/dev/video0
 
−
|/dev/v4l-subdev2
 
−
|}- CM5
 
−
{| class="wikitable"
 
−
!CAM num
 
−
!I2C
 
−
!media node
 
−
!media entity name
 
−
!video node
 
−
!subdev node
 
−
|-
 
−
|1
 
−
|4
 
−
|/dev/media2
 
−
|m00_b_mvcam 4-003b
 
−
|/dev/video22
 
−
|/dev/v4l-subdev8
 
−
|-
 
−
|2
 
−
|3
 
−
|/dev/media3
 
−
|m01_b_mvcam 3-003b
 
−
|/dev/video33
 
−
|/dev/v4l-subdev11
 
−
|-
 
−
|3
 
−
|5
 
−
|/dev/media1
 
−
|m00_b_mvcam 5-003b
 
−
|/dev/video11
 
−
|/dev/v4l-subdev5
 
−
|-
 
−
|4
 
−
|6
 
−
|/dev/media0
 
−
|m00_b_mvcam 6-003b
 
−
|/dev/video0
 
−
|/dev/v4l-subdev2
 
−
|}
 
−
====相机支持的mbus-code====
 
−
MV系列和RAW系列相机具备不同的数据格式能力，具体请参考每个型号的相机的数据手册。
 
−
{| class="wikitable"
 
−
|+格式对应关系
 
−
!Format on datasheet
 
−
!mbus-code for media-ctl
 
−
!FourCC pixelformat for v4l2-ctl
 
−
|-
 
−
|RAW8
 
−
|Y8_1X8
 
−
|GREY
 
−
|-
 
−
|RAW10
 
−
|Y10_1X10
 
−
|'Y10 '
 
−
|-
 
−
|RAW12
 
−
|Y12_1X12
 
−
|'Y12 '
 
−
|-
 
−
|UYVY
 
−
|UYVY8_2X8
 
−
|UYVY
 
−
|}
 
−
===raw图片格式说明===
 
−
RK35xx的VICAP模块支持两种数据保存格式，Compact和Noncompact RAW。可以通过RKCIF的RKCIF_CMD_SET_CSI_MEMORY_MODE ioctl指令修改模式。默认情况下，输出的是Compact RAW格式。[[File:Compact raw and noncompact raw of rk3588 vicap.png|center|thumb|800x800px|Compact raw and noncompact raw of rk3588 VICAP|link=http://wiki.veye.cc/index.php/File:Compact_raw_and_noncompact_raw_of_rk3588_vicap.png]]<br />
 
−
====具有填充位的图片格式（Noncompact RAW）====
 
−
对于10bit深度、12bit深度的像素数据，总是使用两个字节来存储一个像素。这种存储方式方便软件处理，缺点是空间占用较大。根据数据有效数据保存在高位还是低位，又分为high align和low align两种。
 
−
=====Noncompact RAW(high align)=====
 
−
数据保存到高位，用不到的低位以0填充。RK VICAP支持的一种数据格式。
 
−
=====Noncompact RAW(low align)=====
 
−
数据保存到低位，用不到的高位以0填充。V4L2标准的'Y10 ' (10-bit Greyscale),'Y12 ' (12-bit Greyscale)格式都是以这种方式存储的。
 
−
 
 
−
后文提到的pixel_layer_convert转换工具，也是将Compact RAW转换为此种存储方式。方便使用图片播放器打开显示。
 
−
====紧密排列的图片格式（Compact RAW）====
 
−
如上图所示，像素和像素之间不会进行空bit位填充。
 
−
====行对齐（line stride）====
 
−
为了方便对图像进行快速操作，系统通常为每行数据提供行对齐的buffer大小。RK3588使用256字节对齐。
 
−
 
 
−
line_stride = ALIGN_UP（image_width*bits_per_pixel/8,256）
 
−
 
 
−
比如当图像宽度是1456时：
 
−
 
 
−
8bit depth，line_stride=1536。
 
−
 
 
−
10bit depth，preferred_stride=2048。
 
−
 
 
−
12bit depth，preferred_stride=2304。
 
−
====转换工具====
 
−
我们编写了一个小工具：[https://github.com/veyeimaging/pixel_layer_convert pixel_layer_convert]，可以很方便的将紧密排列的图片（Compact RAW）转为具有填充位的图片(Noncompact RAW(low align))。
 
  
−
比如下面的命令，可以将一个1456宽度的compact raw10 图片转为noncompact raw10。
+
<code>mvcam 7-003b: camera is: RAW-MIPI-SC132M</code>
  
−
<code>./pixel_layer_convert -I R10C -i y10-1456x1088_0001.raw -o y10-1456x1088_0001_new.raw -w 1456</code>
+
<code>mvcam 7-003b: firmware version: 0x1040000</code>
−
====raw图片播放器====
 
−
推荐使用[https://www.offminor.de/ vooya]作为播放器，支持GREY, 和具有填充位的图片格式。
 
  
−
Y8文件可以用这个播放器： [https://yuv-player-deluxe.software.informer.com/2.6/ YUV Displayer Deluxe]。
+
===Camera Application Development Guide===
−
===应用范例===
+
[[MV Camera Application Development Guide on Rochchip|Application Development Guide]]
−
请注意，下面章节中的<code>/dev/media0</code>,<code>/dev/video0</code>,<code>/dev/v4l-subdev2</code>根据实际需要替换为前面章节的描述进行替换。
 
  
−
====配置命令行全局变量====
+
===References===
−
根据主板型号，配置I2C_BUS全局变量。
 
  
 
*OrangePi CM4
 
*OrangePi CM4
  
−
<code>export I2C_BUS=1</code>
+
http://www.orangepi.cn/orangepiwiki/index.php/Orange_Pi_CM4
  
 
*OrangePi CM5
 
*OrangePi CM5
  
−
<code>export I2C_BUS=6</code>
+
http://www.orangepi.cn/html/hardWare/computerAndMicrocontrollers/service-and-support/Orange-Pi-CM5.html
−
 
+
===Document History===
−
为了后面描述方便，此处根据sensor尺寸，配置全局变量。
 
−
 
 
−
*MV-MIPI-IMX178M
 
−
 
 
−
<code>export WIDTH=3088</code>
 
−
 
 
−
<code>export HEIGHT=2064</code>
 
−
 
 
−
<code>export FPS=22</code>
 
−
 
 
−
*MV-MIPI-SC130M
 
−
 
 
−
<code>export WIDTH=1280</code>
 
−
 
 
−
<code>export HEIGHT=1024</code>
 
−
 
 
−
<code>export FPS=108</code>
 
−
 
 
−
*MV-MIPI-IMX296M
 
−
 
 
−
<code>export WIDTH=1456</code>
 
−
 
 
−
<code>export HEIGHT=1088</code>
 
−
 
 
−
<code>export FPS=60</code>
 
−
 
 
−
*MV-MIPI-IMX287M
 
−
 
 
−
<code>export WIDTH=704</code>
 
−
 
 
−
<code>export HEIGHT=544</code>
 
−
 
 
−
<code>export FPS=319</code>
 
−
 
 
−
*MV-MIPI-IMX265M
 
−
 
 
−
<code>export WIDTH=2048</code>
 
−
 
 
−
<code>export HEIGHT=1544</code>
 
−
 
 
−
<code>export FPS=45</code>
 
−
 
 
−
*MV-MIPI-IMX264M
 
−
 
 
−
<code>export WIDTH=2432</code>
 
−
 
 
−
<code>export HEIGHT=2056</code>
 
−
 
 
−
<code>export FPS=28</code>
 
−
 
 
−
*RAW-MIPI-SC132M
 
−
 
 
−
<code>export WIDTH=1024</code>
 
−
 
 
−
<code>export HEIGHT=1280</code>
 
−
 
 
−
<code>export FPS=120</code>
 
−
 
 
−
*RAW-MIPI-IMX462M
 
−
 
 
−
<code>export WIDTH=1920</code>
 
−
 
 
−
<code>export HEIGHT=1088</code>
 
−
 
 
−
<code>export FPS=60</code>
 
−
 
 
−
*RAW-MIPI-AR0234M
 
−
 
 
−
<code>export WIDTH=1920</code>
 
−
 
 
−
<code>export HEIGHT=1200</code>
 
−
 
 
−
<code>export FPS=60</code>
 
−
 
 
−
*RAW-MIPI-SC535M
 
−
 
 
−
<code>export WIDTH=2432</code>
 
−
 
 
−
<code>export HEIGHT=2048</code>
 
−
 
 
−
<code>export FPS=50</code>
 
−
====使用v4l2-ctl配置参数====
 
−
<code>$ v4l2-ctl -d /dev/v4l-subdev2 -L</code>
 
−
 
 
−
<code>User Controls</code>
 
−
 
 
−
<code>                   trigger_mode 0x00981901 (int)    : min=0 max=2 step=1 default=0 value=0 flags=volatile, execute-on-write</code>
 
−
 
 
−
<code>                    trigger_src 0x00981902 (int)    : min=0 max=1 step=1 default=1 value=1 flags=volatile, execute-on-write</code>
 
−
 
 
−
<code>                    soft_trgone 0x00981903 (button) : flags=write-only, execute-on-write</code>
 
−
 
 
−
<code>                     frame_rate 0x00981904 (int)    : min=1 max=60 step=1 default=60 value=60 flags=volatile, execute-on-write</code>
 
−
 
 
−
<code>                          roi_x 0x00981905 (int)    : min=0 max=1376 step=8 default=0 value=0</code>
 
−
 
 
−
<code>                          roi_y 0x00981906 (int)    : min=0 max=1024 step=4 default=0 value=0</code>
 
−
 
 
−
可以使用下面的方法设置和获取参数。
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl [ctrl_type]=[val]</code>
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --get-ctrl [ctrl_type]</code>
 
−
 
 
−
以上所有功能，均可使用mv_mipi_i2c.sh实现。
 
−
 
 
−
需要指出的是，以上参数在取图过程中均不可修改。下面进行逐个说明：
 
−
=====配置触发模式=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_mode=[0-2]</small></code>
 
−
 
 
−
0：流模式
 
−
 
 
−
1：普通触发模式
 
−
 
 
−
2：快速连续触发模式
 
−
=====配置触发源=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_src=[0-1]</small></code>
 
−
 
 
−
0: 软触发
 
−
 
 
−
1: 硬触发
 
−
=====软触发一次=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>soft_trgone=1</small></code>
 
−
=====设置帧率=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl frame_rate=[1-max]</code>
 
−
 
 
−
随分辨率的调整，最大帧率会自动更新。
 
−
=====设置ROI起始位置=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_x=0</code>
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_y=0</code>
 
−
 
 
−
ROI其实位置设置之后，需要通过<code>media-ctl</code>命令完成完整的ROI配置。
 
−
 
 
−
设置roi后，最大帧率会自动调整。请注意，摄像机roi参数需要符合摄像机手册中的要求。
 
−
====使用media-ctl配置图像格式====
 
−
可以通过下面的命令，配置摄像机的数据格式，分辨率，帧率。
 
−
 
 
−
<code>media-ctl -d /dev/media0 --set-v4l2 '"m00_b_mvcam '"$I2C_BUS"'-003b":0[fmt:Y8_1X8/'"$WIDTH"'x'"$HEIGHT"'@1/'"$FPS"']'</code>
 
−
 
 
−
其中：<code>"m00_b_mvcam 7-003b"</code>指的相机的<code>entity</code>完整名称，<code>Y8_1X8</code>为<code>mbus-code</code>，<code>'"$WIDTH"'x'"$HEIGHT"'</code>指明了分辨率，<code>1/'"$FPS"'</code>指明了帧率。
 
−
 
 
−
此处的width和height配合v4l2-ctl命令的<code>roi_x</code>和<code>roi_y</code>共同组成了ROI参数。
 
−
 
 
−
比如，对于MV-MIPI-IMX296M,这个命令经过变量替换后的结果是：
 
−
 
 
−
<code>media-ctl -d /dev/media0 --set-v4l2 '"m00_b_mvcam 7-003b":0[fmt:Y8_1X8/1456x1088@1/60 field:none]'</code>
 
−
 
 
−
不仅可以在一个命令中同时配置数据格式，分辨率，帧率，也可以单独修改某一项或某几项。
 
−
====流模式====
 
−
=====设置数据格式，分辨率，帧率=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_x=0</code>
 
  
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_y=0</code>
+
*2025-12-26
  
−
<code>media-ctl -d /dev/media0 --set-v4l2 '"m00_b_mvcam '"$I2C_BUS"'-003b":0[fmt:Y8_1X8/'"$WIDTH"'x'"$HEIGHT"'@1/'"$FPS"']'</code>
+
Add a chapter on application development guidelines
−
=====帧率统计=====
 
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=width=$WIDTH,height=$HEIGHT,pixelformat=GREY --stream-mmap --stream-count=-1 --stream-to=/dev/null</code>
 
  
−
或者
+
*2025-09-16
  
−
<code>./yavta -c-1 --skip 0 -f Y8 -s ${WIDTH}x${HEIGHT} /dev/video0</code>
+
Add support for CM5 tablet.
−
=====保存图片到文件=====
 
  
−
*raw8格式
+
*2025-03-23
  
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=width=$WIDTH,height=$HEIGHT,pixelformat=GREY --stream-mmap --stream-count=1 --stream-to=y8-${WIDTH}x${HEIGHT}.raw</code>
+
Add description of mv_probe.sh.
−
 
 
−
*raw10格式
 
−
 
 
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=width=$WIDTH,height=$HEIGHT,pixelformat='Y10 ' --stream-mmap --stream-count=1 --stream-to=y10-${WIDTH}x${HEIGHT}.raw</code>
 
−
 
 
−
*raw12格式
 
−
 
 
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=$WIDTH,height=$HEIGHT,pixelformat='Y12 ' --stream-mmap --stream-count=1 --stream-to=y12-${WIDTH}x${HEIGHT}.raw</code>
 
−
 
 
−
图片格式请参考前面章节的描述。
 
−
=====yavta应用范例=====
 
−
======yavta安装======
 
−
<code>git clone <nowiki>git://git.ideasonboard.org/yavta.git</nowiki></code>
 
−
 
 
−
<code>cd yavta;make</code>
 
−
======保存图片到文件======
 
−
配置完数据格式，分辨率，帧率之后，执行：
 
−
 
 
−
<code>./yavta -c1 -Fy8-${WIDTH}x${HEIGHT}.raw --skip 0 -f Y8 -s ${WIDTH}x${HEIGHT} /dev/video0</code>
 
−
=====将相机数据导入OpenCV的范例=====
 
−
<code>sudo apt install python3-opencv</code>
 
−
 
 
−
详见github上的[https://github.com/veyeimaging/rk356x_firefly/tree/main/linux/samples samples]目录。
 
−
 
 
−
<code>python3 ./v4l2dev_2_opencv_show_grey.py --width 1456 --height 1088 --fps 60 --i2c 6</code>
 
−
 
 
−
注意需要使用合适的参数来执行上述程序。
 
−
=====gstreamer应用范例=====
 
−
我们提供了几个gstreamer例程，实现了预览功能。详见github上的[https://github.com/veyeimaging/rk35xx_veye_bsp/tree/main/samples samples]目录。
 
−
====触发模式====
 
−
=====设置数据格式，分辨率，帧率=====
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_x=0</code>
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl roi_y=0</code>
 
−
 
 
−
<code>media-ctl -d /dev/media0 --set-v4l2 '"m00_b_mvcam '"$I2C_BUS"'-003b":0[fmt:Y8_1X8/'"$WIDTH"'x'"$HEIGHT"'@1/'"$FPS"']'</code>
 
−
=====软触发模式=====
 
−
======设置模式======
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_mode=1</small></code>
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_src=0</small></code>
 
−
======开始取图======
 
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=width=$WIDTH,height=$HEIGHT,pixelformat=GREY --stream-mmap --stream-count=1 --stream-to=y8-${WIDTH}x${HEIGHT}.raw</code>
 
−
======进行软触发操作======
 
−
在其他的shell终端，可以多次执行下面命令进行多次触发。
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>soft_trgone=1</small></code>
 
−
=====硬触发模式=====
 
−
======设置模式======
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_mode=1</small></code>
 
−
 
 
−
<code>v4l2-ctl -d /dev/v4l-subdev2 --set-ctrl <small>trigger_src=1</small></code>
 
−
 
 
−
可以使用[[Mv mipi i2c.sh user guide|mv_mipi_i2c.sh]]脚本进行丰富的触发参数设置。
 
−
======开始取图======
 
−
<code>v4l2-ctl -d /dev/video0 --set-fmt-video=width=$WIDTH,height=$HEIGHT,pixelformat=GREY --stream-mmap --stream-count=1 --stream-to=y8-${WIDTH}x${HEIGHT}.raw</code>
 
−
======进行硬触发操作======
 
−
将合适的触发信号连接到摄像机的trigger引脚并进行触发。
 
−
===i2c脚本使用说明===
 
−
我们提供了shell脚本来配置参数。
 
−
 
 
−
[[mv_mipi_i2c.sh user guide]]
 
−
===参考资料===
 
−
 
 
−
*OrangePi CM4
 
−
 
 
−
http://www.orangepi.cn/orangepiwiki/index.php/Orange_Pi_CM4
 
−
 
 
−
*OrangePi CM5
 
−
 
 
−
http://www.orangepi.cn/html/hardWare/computerAndMicrocontrollers/service-and-support/Orange-Pi-CM5.html
 
−
===本文修改记录===
 
  
−
*2025-01-09
+
*2024-01-10
  
−
发布第一个版本。
+
Release 1st version.

Latest revision as of 15:42, 26 December 2025

查看中文

1 Overview

The MV series and RAW series cameras are cameras designed for AI applications in the industrial field. They use the MIPI CSI-2 interface and are particularly suitable for use with embedded computing platforms. They have rich data formats and triggering features, extremely low latency, high bandwidth, and reliable stability.

This article takes OrangePi CM4 and OrangePi CM5 and CM5 tablet board as an example to introduce how to connect MV and RAW series cameras to the RK3566/3K3568 and RK3588S/RK3588 system.

We provide drivers for the Linux operating system (using Ubuntu as an example).

1.1 Camera Module List

Series Model Status
MV series MV-MIPI-IMX178M Done
MV series MV-MIPI-SC130M Done
MV series MV-MIPI-IMX296M Done
MV series MV-MIPI-IMX287M Done
MV series MV-MIPI-IMX265M Done
MV series MV-MIPI-IMX264M Done
MV series MV-MIPI-GMAX4002M Done
RAW series RAW-MIPI-SC132M Done
RAW series RAW-MIPI-IMX462M Done
RAW series RAW-MIPI-AR0234M Done
RAW series RAW-MIPI-SC535M Done

2 Hardware Setup

We use the official baseboards of the Orange Pi CM4 and CM5, which feature a 15-pin header compatible with Raspberry Pi. For the RAW series cameras, our cameras can be directly mounted onto the baseboard without the need for an adapter board. For the MV series cameras, the ADP-MV1 adapter board is required for connection.

2.1 Camera Connection to Orange Pi CM4

The ADP-MV1 is connected to the OrangePi CM4 via a 15-pin FFC cable with opposite surface alignment; please pay attention to the orientation of the contact surfaces.

Note that only the CAM1 shown in the image below supports MV and RAW cameras.

OrangePi CM4 to MV cam
OrangePi CM4 to RAW cam


2.2 Camera Connection to Orange Pi CM5

The OrangePi CM5 supports up to four cameras. The following diagram shows the hardware connection method for simultaneously connecting multiple cameras.

OrangePi CM5 to all cam overview
OrangePi CM5 to all cam backview


2.3 Camera Connection to Orange Pi CM5 tablet

Due to differences in connector formats, we have successfully interfaced modules for both CAM2 and CAM3 on the OrangePi CM5 tablet. The diagram below illustrates the connection methods for the RAW series and MV series.

OrangePi CM5 tablet to RAW cam
OrangePi CM5 tablet to MV cam

3 Introduction to github repositories

https://github.com/veyeimaging/rk35xx_veye_bsp

https://github.com/veyeimaging/rk35xx_orangepi

includes：

  • driver source code
  • i2c toolkits
  • application demo

In addition, a compiled linux kernel installation package is provided in the releases.

4 Upgrade the Ubuntu system

We provide a flashing image for the release system, as well as a deb package for the Linux kernel.

Refer to the OrangePi CM4 user manual or the OrangePi CM5 user manual for instructions on flashing the system. Alternatively, you can use the general dpkg command to install the deb package.

5 Check system status

Run the following command to confirm whether the camera is probed.

sudo dmesg | grep mvcam

5.1 CM4

The CM4 supports camera connection only through the CAM1 interface. Taking the RAW-MIPI-SC132M as an example, the dmesg output contains the following information:

The output message appears as shown below：

mvcam 1-003b: camera is: RAW-MIPI-SC132M

mvcam 1-003b: firmware version: 0x1040000

  • Run the following command to check the presence of video node.

ls /dev/video0

The output message appears as shown below.

video0

5.2 CM5

The CM5 supports the connection of up to four cameras. Taking the RAW-MIPI-SC132M as an example, the dmesg output contains the following information:

mvcam 3-003b: camera is: RAW-MIPI-SC132M

mvcam 3-003b: firmware version: 0x1040000

mvcam 4-003b: camera is: RAW-MIPI-SC132M

mvcam 4-003b: firmware version: 0x1040000

mvcam 5-003b: camera is: RAW-MIPI-SC132M

mvcam 5-003b: firmware version: 0x1040000  

mvcam 6-003b: camera is: RAW-MIPI-SC132M

mvcam 6-003b: firmware version: 0x1040000

5.3 CM5 tablet

The CM5 supports the connection of up to two cameras. Taking the RAW-MIPI-SC132M as an example, the dmesg output contains the following information:

mvcam 6-003b: camera is: RAW-MIPI-SC132M

mvcam 6-003b: firmware version: 0x1040000

mvcam 7-003b: camera is: RAW-MIPI-SC132M

mvcam 7-003b: firmware version: 0x1040000

6 Camera Application Development Guide

Application Development Guide

7 References

  • OrangePi CM4

http://www.orangepi.cn/orangepiwiki/index.php/Orange_Pi_CM4

  • OrangePi CM5

http://www.orangepi.cn/html/hardWare/computerAndMicrocontrollers/service-and-support/Orange-Pi-CM5.html

8 Document History

  • 2025-12-26

Add a chapter on application development guidelines

  • 2025-09-16

Add support for CM5 tablet.

  • 2025-03-23

Add description of mv_probe.sh.

  • 2024-01-10

Release 1st version.

Retrieved from "http://wiki.veye.cc/index.php?title=MV_Camera_on_Orange_Pi%27s_RK35XX_Boards&oldid=8971"