Intel renascence working issue
hi,
I am currently utilizing the Intel-renascence D410 on Raspberry Pi 4B, which has 4GB of RAM and runs on Raspbian GNU/Linux 10 (buster). Additionally, I am using SDK 2.50.0 and the intel-renascence firmware versions 5.12 and 5.13. My application involves capturing depth frames with the RS2_STREAM_DEPTH configuration, specifically at a resolution of 480X270 and in RS2_FORMAT_Z16 format, with a frame rate of 30 FPS. I am encountering issues with missing frames and USB disconnections when using both a single sensor and a dual sensor. After updating to firmware version 5.14, the USB disconnect issue has been reduced, but the problem of missing frames persists. Furthermore, the issue with dual sensor usage appears to be more pronounced compared to using a single sensor. I kindly request your suggestions to address these challenges.
Regards
Chandiran.S
-
Hi Chandiran When you mention 'dual sensor', do you mean two separate cameras (the D410 and another camera). If so then Raspberry Pi single-board computers usually have only one RealSense camera attached per Pi board.
Other single-board computer types such as Nvidia Jetson and Up Squared are better suited to having more than one camera attached and operating simultaneously.
-
Dear MartyG,Thanks for the information.We are using RPI 4 Model B series of SBC which has 4G/8GB RAM, 04 USB Port (2.0 & 3.0), Gigabit Ethernet, also has package of H.265 decoder, 1080p30 encoder, Open GL ES 3.1 , Vulkan 1.0.Linux kernel : 5.10.63-v7l+Intel SDK compatible version:Moreover our end application requires only to capture IR & depth frames with 30FPS from 02 sensors.Kindly suggest what are the recommended specs for COTS/Custom SBC to work with 02 Intelsensor simultaneously connected to single SBC.Regardschandiran.S
-
Aside from the mentioned Nvidia Jetson and Up SBC boards that are known to work with two RealSense cameras (and also Intel NUC computing products), it is difficult to provide recommendations for specific board products that can support multiple cameras.
A general rule of thumb that I use for specifications is capability equal to Intel Core i3 computers for 2 simultaneously active cameras, Core i5 for 3 cameras and Core i7 / i9 for 4 or more cameras.
-
Dear MartyG,Thanks for the information.
The Raspberry Pi 4 B (Broadcom BCM2711) processor has similar performance benchmarks to the Intel Core i3, as shown in the comparison links provided.(https://www.cpu-monkey.com/en/compare_cpu-raspberry_pi_4_b_broadcom_bcm2711-vs-intel_core_i3_4130 ) and (https://cpu-compare.com/cpu/compare/raspberry_pi_4_b_broadcom_bcm2711-vs-intel_core_i3-4160t).
Therefore, like the Intel Core i3, the Raspberry Pi 4 B should be capable of connecting 2 active cameras. Moreover, when running 2 cameras, the Raspberry Pi 4 B has an average CPU load of 30-40%, indicating it has sufficient resources to handle dual camera streaming.
Regards
Chandiran.S
Please sign in to leave a comment.
Comments
5 comments