ON Semiconductor AR0820 Linux Driver

Problems running the pipelines shown on this page? Please see our GStreamer Debugging guide for help.

Omnivision AR0820 Features

The AR0820AT is a 1/2−inch CMOS digital image sensor with a 3848 H x 2168 V active−pixel array. This advanced automotive sensor captures images in either linear or high dynamic range, with rolling−shutter readout.

AR0820AT is optimized for both low light and challenging high dynamic range scene performance, with a 2.1 µm DR−Pix BSI pixel and on−sensor 140 dB HDR capture capability. The sensor includes advanced functions such as in−pixel binning, windowing, and both video and single frame modes to provide flexible Region of Interest (ROI) or specific resolution in order to enhance performance in extreme low light conditions. The sophisticated sensor fault detection features and embedded data on AR0820AT are designed to enable camera ASIL B compliance. The device is programmable through a simple two−wire serial interface and supports MIPI output interface.

Supported Platforms

• NVIDIA Jetson Xavier

Features Included in the Driver

Xavier	Feature Details SDK Support 3848x2168@30fps GMSL2, 4 Lanes, RAW12 L4T 32.5.1 / Jetpack 4.5.1

Enabling the Driver

In order to use this driver, you have to patch and compile the kernel source using JetPack:

• Follow the instructions to get the kernel source code.

• Once you have the source code, apply the following the patches in order to add the changes required for the IMX327 camera at kernel and dtb level.

4.5.1_ar0820.patch

• Follow the instructions for building the kernel, and then flash the image.

Make sure to enable IMX327 driver support:

make menuconfig

-> Device Drivers                                                                                                                                                                         
   -> Multimedia support (MEDIA_SUPPORT [=y])                                                                                                                                                            
        -> NVIDIA overlay Encoders, decoders, sensors and other helper chips                                                                                                                       
            -> MAX9296 Deserializer I2C IO Expander (I2C_IOEXPANDER_DESER_MAX9296 [=y]) 

And to select the runtime device tree blob by editing the $JETSON_L4T/rootfs/boot/extlinux/extlinux.conf to add the "FDT" line:

TIMEOUT 30
DEFAULT primary

MENU TITLE L4T boot options

LABEL primary
      MENU LABEL primary kernel
      LINUX /boot/Image
      INITRD /boot/initrd
      FDT /boot/tegra210-p3448-0000-p3449-0000-a02.dtb
      APPEND ${cbootargs} quiet

Using the Driver

GStreamer Examples

Capture and Display

• 1920x1080@30fps RGGB12

gst-launch-1.0 nvarguscamerasrc ! 'video/x-raw(memory:NVMM), width=1920, height=1080, format=NV12, framerate=30/1' ! nvvidconv ! xvimagesink

Video Encoding

CAPS="video/x-raw(memory:NVMM), width=(int)1920, height=(int)1080, format=(string)NV12, framerate=(fraction)30/1"

gst-launch-1.0 nvarguscamerasrc sensor-id=0 num-buffers=500 ! "video/x-raw(memory:NVMM), width=(int)1920, height=(int)1080, format=(string)NV12, framerate=(fraction)30/1" ! omxh264enc ! mpegtsmux ! filesink location=test.ts

The sensor will capture in the 1920x1080@30fps mode and the pipeline will encode the video and save it into test.ts file.

Performance

ARM Load

Tegrastats display the following output when capturing with the sensor driver used in the Jetson Nano platform:

RAM 1167/3963MB (lfb 522x4MB) CPU [25%@1132,16%@1132,9%@1132,12%@1132]
RAM 1168/3963MB (lfb 522x4MB) CPU [28%@921,12%@921,9%@921,13%@921]
RAM 1167/3963MB (lfb 522x4MB) CPU [23%@921,12%@921,13%@921,10%@921]
RAM 1167/3963MB (lfb 522x4MB) CPU [28%@921,8%@921,12%@921,12%@921]
RAM 1169/3963MB (lfb 522x4MB) CPU [26%@1479,9%@1479,16%@1479,9%@1479]
RAM 1167/3963MB (lfb 522x4MB) CPU [28%@921,13%@921,9%@921,16%@921]
RAM 1168/3963MB (lfb 522x4MB) CPU [23%@1036,13%@1036,14%@1036,7%@1036]
RAM 1167/3963MB (lfb 522x4MB) CPU [25%@921,12%@921,9%@921,11%@921]
RAM 1168/3963MB (lfb 522x4MB) CPU [25%@921,13%@921,16%@921,12%@921]
RAM 1169/3963MB (lfb 522x4MB) CPU [27%@921,12%@921,8%@921,13%@921]
RAM 1168/3963MB (lfb 522x4MB) CPU [24%@921,8%@921,13%@921,10%@921]
RAM 1169/3963MB (lfb 522x4MB) CPU [29%@921,13%@921,15%@921,6%@921]

Framerate

Using the next pipeline we were able to measure the framerate for single capture with perf element:

gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! 'video/x-raw(memory:NVMM), width=(int)1920, height=(int)1080, format=(string)NV12, framerate=(fraction)30/1' ! perf  ! fakesink

GST-PERF-INFO --> timestamp: 0:44:34.324884537; bps: 0,000; mean_bps: 0,000; fps: 0,000; mean_fps: 0,000
GST-PERF-INFO -->  timestamp: 0:44:35.354956530; bps: 24192,000; mean_bps: 8064,000; fps: 30,095; mean_fps: 30,095
GST-PERF-INFO -->  timestamp: 0:44:36.355520992; bps: 241920,000; mean_bps: 66528,000; fps: 29,983; mean_fps: 30,039
GST-PERF-INFO -->  timestamp: 0:44:37.356864989; bps: 241920,000; mean_bps: 101606,400; fps: 29,960; mean_fps: 30,013
GST-PERF-INFO -->  timestamp: 0:44:38.357433006; bps: 241920,000; mean_bps: 124992,000; fps: 29,983; mean_fps: 30,005
GST-PERF-INFO -->  timestamp: 0:44:39.358908010; bps: 241920,000; mean_bps: 141696,000; fps: 29,956; mean_fps: 29,995
GST-PERF-INFO -->  timestamp: 0:44:40.359357860; bps: 241920,000; mean_bps: 154224,000; fps: 29,987; mean_fps: 29,994
GST-PERF-INFO -->  timestamp: 0:44:41.360617558; bps: 241920,000; mean_bps: 163968,000; fps: 29,962; mean_fps: 29,989
GST-PERF-INFO -->  timestamp: 0:44:42.361400607; bps: 241920,000; mean_bps: 171763,200; fps: 29,977; mean_fps: 29,988
GST-PERF-INFO -->  timestamp: 0:44:43.362674329; bps: 241920,000; mean_bps: 178141,091; fps: 29,962; mean_fps: 29,985
GST-PERF-INFO -->  timestamp: 0:44:44.363320878; bps: 241920,000; mean_bps: 183456,000; fps: 29,981; mean_fps: 29,984
GST-PERF-INFO -->  timestamp: 0:44:45.364541434; bps: 241920,000; mean_bps: 187953,231; fps: 29,963; mean_fps: 29,983
GST-PERF-INFO -->  timestamp: 0:44:46.365041950; bps: 241920,000; mean_bps: 191808,000; fps: 29,985; mean_fps: 29,983
GST-PERF-INFO -->  timestamp: 0:44:47.366186373; bps: 241920,000; mean_bps: 195148,800; fps: 29,966; mean_fps: 29,981
GST-PERF-INFO -->  timestamp: 0:44:48.366852845; bps: 241920,000; mean_bps: 198072,000; fps: 29,980; mean_fps: 29,981
GST-PERF-INFO -->  timestamp: 0:44:49.368081920; bps: 241920,000; mean_bps: 200651,294; fps: 29,963; mean_fps: 29,980
GST-PERF-INFO -->  timestamp: 0:44:50.368731947; bps: 241920,000; mean_bps: 202944,000; fps: 29,981; mean_fps: 29,980
GST-PERF-INFO -->  timestamp: 0:44:51.370037391; bps: 241920,000; mean_bps: 204995,368; fps: 29,961; mean_fps: 29,979
GST-PERF-INFO -->  timestamp: 0:44:52.370821395; bps: 241920,000; mean_bps: 206841,600; fps: 29,976; mean_fps: 29,979
GST-PERF-INFO -->  timestamp: 0:44:53.371545430; bps: 241920,000; mean_bps: 208512,000; fps: 29,978; mean_fps: 29,979
GST-PERF-INFO -->  timestamp: 0:44:54.372675500; bps: 241920,000; mean_bps: 210030,545; fps: 29,966; mean_fps: 29,978
GST-PERF-INFO -->  timestamp: 0:44:55.373703465; bps: 241920,000; mean_bps: 211417,043; fps: 29,969; mean_fps: 29,978

RidgeRun Resources

Quick Start Client Engagement Process RidgeRun Blog Homepage Technical and Sales Support RidgeRun Online Store RidgeRun Videos Contact Us

Contact Us Visit our Main Website for the RidgeRun Products and Online Store. RidgeRun Engineering informations are available in RidgeRun Professional Services, RidgeRun Subscription Model and Client Engagement Process wiki pages. Please email to support@ridgerun.com for technical questions and contactus@ridgerun.com for other queries. Contact details for sponsoring the RidgeRun GStreamer projects are available in Sponsor Projects page.