I am a software engineer of the VEYE team and will try my best to help you.
I was thinking that since you took it from a dual board down to a single board that you may be able to reduce the board size and break it out into 2 again to fit everything.
Yes, we have the ability. However, the decision to make a new version of model is more of a commercial consideration. We are now more focused on developing new sensor and new products.
I will seriously discuss your suggestion with my colleagues in the company.
Thank you for your approval and feedback.
@radiohound As far as I know, sony's datasheet is only open to corporate customers who have signed an NDA.
@tokabe
Hi,
On i.MX8m Platform, our roadmap is based on SOM board. For now, we have finished MYIR i.MX8m board and AVNET Maaxboard(mini) board.
http://wiki.veye.cc/index.php/NXP_i.MX_Application_Note_Index
The coral.ai board from google is in progress.
So the linux version we use now depends on the system version that comes with the SOM board.
We have a V5.y version driver on raspiberry PI, maybe this will help you.
http://wiki.veye.cc/index.php/V4L2_mode_for_Raspberry_Pi
Of course, we will transplant all product drivers, including CS-MIP-IMX307, to i.MX8m. This work is also in progress.
@desperex said in Please advise compatible mono camera:
About two or three years ago, we tried to connect VEYE series products to rk3399, but at that time there were some problems with its BSP.
Recently we restarted the development plan for rockchip chips.
The first thing we are doing is to connect VEYE series to firefly's RK3566 board. Now the development has been completed and we have made http://wiki.veye.cc/index.php/ADP-Tfirefly. This board is already online for sale.
Next, we have to create the adapter board for MV series to ROC-RK3566/8-PC and develop it.
As for the two boards you mentioned, the rk3399 chip is too old and we have no plans for now.
Orange Pi 5, we have not studied it before. I will submit it to our R&D department.
@lubertroft
hi,
http://wiki.veye.cc/index.php/CS-MIPI-SC132_version_log
yuyv is supported in this version .
Because it is CS series, of course cs_mipi_i2c.sh is the corresponding script.
http://wiki.veye.cc/index.php/VEYE-MIPI-290/327_i2c/#yuvseq
@consti10 said in Improve DE-coding on snapdragon devices:
turning off the HW decoder for h264 on android doesn't really make sense.
Maybe changing the poc type could fix this issue, too ?
I'm not sure about that.
- When connected via USB (original cable) there can be artifacts (I suspect packet loss via usb or some driver bug)
You are right. This camera will have a probability of packet loss. The higher the cpu performance of the host, the less packet loss.
This problem is still being studied and solved.
- POC type as described above.
@esjeeybi
About how to compile our drivers on rpi:
http://wiki.veye.cc/index.php/V4L2_mode_for_Raspberry_Pi#Source_code_cross_compilation
You can get/set FPS, Gain, Brightness, Contrast via:
http://wiki.veye.cc/index.php/CS-MIPI-X_i2c
@ess
Hi, thank you, it is indeed typo.
jetson supports trigger mode, which I have done research on and it has worked. The plan is to organize it and release it along with our new machine vision product line. Estimated 15 days or so.
@ess
Our company discussed and decided to bring this forward. It is expected that the version will be available to you within 10 days.
@lkarthus 你好，现在307这款的锐度调节是自动进行的，没有提供外部设置接口。
@chjchoi
Is your tx2 the official Nvidia devkit?
Did you use our direct compiled dtb and Image, or did you recompile the code yourself? Did you make any changes to the code?
@powerliusu
你的这几个操作并没有必然逻辑的相关性。
@powerliusu said in jetson nano 在移植核后无法再烧录镜像:
apt-mark hold
这个的作用是标记一下防止后面apt upgrade的时候把kernel也升级掉。本身不是必须的步骤，你可以先忽略掉它。
@powerliusu said in jetson nano 在移植核后无法再烧录镜像:
i2cdetect 总线1的10寄存器数值改了
这个命令的作用是检测某个i2cbus上存在哪些i2c设备，你用的是nanoa02,对应的i2cbus应该是6。 替换完成Image和dtb之后，重启，然后用i2cdetect -y -r 6作为辅助调试手段即可。
至于sd卡无法烧录，建议你格式化这个tf卡后再试一次，或者换个tf卡试一下。
整体上而言，步骤是：烧写标准系统-->更新Image和dtb-->系统重启-->检测状态。
@milank I suspect that the FFC cable is not connected well enough. You could try replacing the FFC cable with a new one, or re-plugging it.
@desperex said in Please advise compatible mono camera:
EDIT: There's also a warning "mvcam 4-003b: Consider updating driver mvcam to match on endpoints", not sure if it's harmful or not
This will not cause problems.
For machine vision type cameras, many times the isp function is not needed. Many steps in isp are dealing with color, and this is certainly not necessary to do. There are also valid operations such as AE, gamma, etc. But there are many customer application scenarios that do not need these functions.
Which functions of isp do you need, please?
Hello, for rockchip platform, we are now accessing the chip rk3566 RK3568 RK3588, etc. For the time being, there is no plan to access to rk3599.
The motherboard manufacturer of our choice is firefly, and provides the corresponding adapter board.
For details, you can refer to the following link. The driver on the rk35XX, which will help you to make the docking.
http://wiki.veye.cc/index.php/VEYE_CS_Camera_on_Firfly_Boards
@nathan-sinclair said in using veye-mipi-327, unedited h264 video for display in html video element:
Just a query about whether it's possible to be sending the h264 video from the mipi-327 camera unedited, so that it can be displayed in a video element on a web page.
The way we are currently doing this, is using a library on the client side called jmuxer, which, through a websocket directly is able to display arducam cameras. Is there something similar with the mipi-327 camera as seen in libcamera-examples?
Sorry, we now only have a similar demo available in legacy mode via gstreamer.
http://wiki.veye.cc/index.php/VEYE-MIPI-290/327_for_Raspberry_Pi#veye_raspcam:_raspcam_alike_toolkits
@milank
Are you using it on a Raspberry Pi? Can you tell me about the software and hardware configuration of your board?
@thn The possible reason is that there are other signals on the i2c bus, causing the camera to misjudge.
Is there any other device mounted on your i2c bus?
