我在NVIDIA Jetson TX2上的Ubuntu 18.04上使用Python 3.6。所有软件都是最新的。 Linux内核是4.9.140-tegra。 GStreamer是版本1.14.1。
我有一个UVC Capable网络摄像头,它显示以下功能,我需要使用Y16(GRAY16_LE)格式进行流式传输:
root@nvidia:/tmp# v4l2-ctl --list-formats-ext -d /dev/video1
ioctl: VIDIOC_ENUM_FMT
Index : 0
Type : Video Capture
Pixel Format: 'UYVY'
Name : UYVY 4:2:2
Size: Discrete 160x120
Interval: Discrete 0.111s (9.000 fps)
Index : 1
Type : Video Capture
Pixel Format: 'Y16 '
Name : 16-bit Greyscale
Size: Discrete 160x120
Interval: Discrete 0.111s (9.000 fps)
Size: Discrete 160x122
Interval: Discrete 0.111s (9.000 fps)
Index : 2
Type : Video Capture
Pixel Format: 'GREY'
Name : 8-bit Greyscale
Size: Discrete 160x120
Interval: Discrete 0.111s (9.000 fps)
Index : 3
Type : Video Capture
Pixel Format: 'RGBP'
Name : 16-bit RGB 5-6-5
Size: Discrete 160x120
Interval: Discrete 0.111s (9.000 fps)
Index : 4
Type : Video Capture
Pixel Format: 'BGR3'
Name : 24-bit BGR 8-8-8
Size: Discrete 160x120
Interval: Discrete 0.111s (9.000 fps)
使用命令行中的GStreamer,我可以使用gst-launch以我想要的格式成功地使设备进入PLAYING状态:
root@nvidia:/tmp# gst-launch-1.0 v4l2src device=/dev/video1 ! video/x-raw,width=160,height=122,format='GRAY16_LE',framerate=9/1 ! fakesink
Setting pipeline to PAUSED ...
Pipeline is live and does not need PREROLL ...
Setting pipeline to PLAYING ...
New clock: GstSystemClock
我试图用Python OpenCV打开GRAY16_LE格式的设备:
import cv2
cap = cv2.VideoCapture("v4l2src device=/dev/video1 ! video/x-raw, format=(string)GRAY16_LE, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1 ! appsink")
ret, frame = cap.read()
在升级的GStreamer调试期间,将在下面探测和打印功能:
0:00:00.260383826 15768 0x49a9850 INFO v4l2 gstv4l2object.c:4136:gst_v4l2_object_probe_caps:<v4l2src0:src> probed caps: video/x-raw, format=(string)UYVY, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1; video/x-raw, format=(string)BGR, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1; video/x-raw, format=(string)RGB16, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1; video/x-raw, format=(string)GRAY8, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1; video/x-raw, format=(string)GRAY16_LE, width=(int)160, height=(int)122, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1; video/x-raw, format=(string)GRAY16_LE, width=(int)160, height=(int)120, pixel-aspect-ratio=(fraction)1/1, framerate=(fraction)9/1
启用GStreamer调试后,错误如下所示:
0:00:00.260385266 15768 0x4744c00 INFO GST_STATES gstelement.c:2579:_priv_gst_element_state_changed:<v4l2src0> notifying about state-changed PAUSED to PLAYING (VOID_PENDING pending)
0:00:00.260529393 15768 0x49a9850 WARN basesrc gstbasesrc.c:3055:gst_base_src_loop:<v4l2src0> error: Internal data stream error.
0:00:00.260531697 15768 0x4744c00 INFO GST_STATES gstbin.c:2952:gst_bin_change_state_func:<pipeline0> child 'v4l2src0' changed state to 4(PLAYING) successfully
0:00:00.260589009 15768 0x49a9850 WARN basesrc gstbasesrc.c:3055:gst_base_src_loop:<v4l2src0> error: streaming stopped, reason not-negotiated (-4)
0:00:00.260660881 15768 0x4744c00 INFO GST_STATES gstbin.c:2087:gst_bin_get_state_func:<pipeline0> getting state
0:00:00.260748913 15768 0x49a9850 INFO GST_ERROR_SYSTEM gstelement.c:2145:gst_element_message_full_with_details:<v4l2src0> posting message: Internal data stream error.
0:00:00.260762833 15768 0x4744c00 INFO GST_STATES gstelement.c:2392:gst_element_get_state_func:<pipeline0> waiting for element to commit state
0:00:00.260909520 15768 0x49a9850 INFO GST_ERROR_SYSTEM gstelement.c:2172:gst_element_message_full_with_details:<v4l2src0> posted error message: Internal data stream error.
0:00:00.261064816 15768 0x4744c00 INFO GST_STATES gstbin.c:2504:gst_bin_element_set_state:<appsink0> current PAUSED pending VOID_PENDING, desired next READY
0:00:00.261189072 15768 0x49a9850 INFO task gsttask.c:316:gst_task_func:<v4l2src0:src> Task going to paused
我使用的是探针返回的相同格式,但错误表示“格式协商错误”。我该如何解决这个问题,以便我的框架出现在Python中?谢谢。
我不得不在OpenCV 4源代码中修补cap_gstreamer.cpp
文件,然后再次构建并安装OpenCV。
那时,我可以用appsink
格式的GRAY16_LE
捕获帧。