Beaglebone Images, Video and OpenCV

Hi Derek,

Thanks for this and the great video. Excellent as always…
I tried to replicate your tutorial, but using the Creative Labs VF0610 Live HD Cam.

…
root@bbb-es-02:~/boneCV# lsusb
Bus 001 Device 002: ID 041e:4080 Creative Technology, Ltd
…
root@bbb-es-02:~/boneCV# v4l2-ctl –all
Driver Info (not using libv4l2):
Driver name : uvcvideo
Card type : VF0610 Live! Cam Socialize HD
…
Video input : 0 (Camera 1: ok)
Format Video Capture:
Width/Height : 960/544
Pixel Format : ‘YUYV’
…

I modified the cpature.c file to use the 960/544 values for width/height.
When I run it, I just see 1 “.” and a “Select timeout” error:
…
root@bbb-es-02:~/boneCV# ./capture -F -o -c 300 > output.raw
Force Format 2
.select timeout
…

There is a raw file, but I thing it is bogus:
…
-rw-r–r– 1 root root 1044480 May 26 00:25 output.raw
…
When I run raw2mpg4, It gives an error too:
…
(see Changelog for the list of incompatible changes).
[h264 @ 0x37a20] missing picture in access unit
[h264 @ 0x37a20] no frame!
[h264 @ 0x2da80] Could not find codec parameters (Video: h264)
[h264 @ 0x2da80] Estimating duration from bitrate, this may be inaccurate
output.raw: could not find codec parameters
…

Any clue what may be wrong here…
thanks,
/venkat

Really nice code display Derek. Is that a WordPress plugin?

I’m getting this error on execution of capture:

“error while loading shared libraries: libv4l2.so.0: cannot open shared object file: No such file or directory”

Installing the following did not fix the error:
apt-get install v4l-utils v4l-conf libav-tools
apt-get install libv4l-dev

Thank you so much for this tutorial code and video, its saved me a ton of time. I modified capture.c to add motion jpeg following your h264 changes:

if (force_format==3){ /* HP 3110 webcam */
fmt.fmt.pix.width = 1280;
fmt.fmt.pix.height = 720;
fmt.fmt.pix.pixelformat = V4L2_PIX_FMT_MJPEG;
fmt.fmt.pix.field = V4L2_FIELD_INTERLACED;
}
if (force_format==2){
fmt.fmt.pix.width = 1920;
fmt.fmt.pix.height = 1080;
fmt.fmt.pix.pixelformat = V4L2_PIX_FMT_H264;
fmt.fmt.pix.field = V4L2_FIELD_INTERLACED;
}

static void usage(FILE *fp, int argc, char **argv)
{
fprintf(fp,
“Usage: %s [options]\n\n”
“Version 1.3\n”
“Options:\n”
“-d | –device name Video device name [%s]\n”
“-h | –help Print this message\n”
“-m | –mmap Use memory mapped buffers [default]\n”
“-r | –read Use read() calls\n”
“-u | –userp Use application allocated buffers\n”
“-o | –output Outputs stream to stdout\n”
“-f | –format Force format to 640×480 YUYV\n”
“-F | –formatH264 Force format to 1920×1080 H264\n”
“-M | –formatMJPEG Force format to 1280×720 MJPEG\n”
“-c | –count Number of frames to grab [%i] – use 0 for infinite\n”
“\n”
“Example usage: capture -F -o -c 300 > output.raw\n”
“Captures 300 frames of H264 at 1920×1080 – use raw2mpg4 script to convert to mpg4\n”,
argv[0], dev_name, frame_count);
}

static const char short_options[] = “d:hmruofFMc:”;

static const struct option
long_options[] = {
{ “device”, required_argument, NULL, ‘d’ },
{ “help”, no_argument, NULL, ‘h’ },
{ “mmap”, no_argument, NULL, ‘m’ },
{ “read”, no_argument, NULL, ‘r’ },
{ “userp”, no_argument, NULL, ‘u’ },
{ “output”, no_argument, NULL, ‘o’ },
{ “format”, no_argument, NULL, ‘f’ },
{ “formatH264”, no_argument, NULL, ‘F’ },
{ “formatMJPEG, no_argument, NULL, ‘M” },
{ “count”, required_argument, NULL, ‘c’ },
{ 0, 0, 0, 0 }
};

In the main() getopts switch:
case ‘F’:
force_format=2;
break;
case ‘M’:
force_format=3;
break;

I hope this can help someone get started using a cheap webcam that only has mjpeg instead of h264 like the HP 3110.

FYI I’m using the BeagleBone “white” Rev A6 with the latest Angstrom image I could find:
http://downloads.angstrom-distribution.org/demo/beaglebone/testing/Angstrom-Cloud9-IDE-GNOME-eglibc-ipk-v2013.06-beaglebone-2013.12.13.img.xz

Haven’t got grabber.c working yet

Hi all…I want to perform Image processing tasks in BBB which runs Debian Linux… so I need to install OpenCV on BBB.. It seems i should install cmake in prior… I dont have direct internet connection to my BBB.. I tried to download OpenCV from github.com as a ZIP file and i copied it to BBB using pendrive….Can anyone please guide me how to install cmake on BBB?? i tried to download cmake and ended up in incomplete build…. please help me …. your ideas will be really appreciated

Experts please help the beginners 🙂

Thanks in advance 🙂

Sir,
I am getting an error “select timeout” while running capture script. I am using a different camera with pixel format of “YUYV” and I have changed the force format option to “-f” according to that but the error is not going away. There is some difference in the drivers of camera also. Following are the full details of camera.

root@beaglebone:~/boneCV# v4l2-ctl –all
Driver Info (not using libv4l2):
Driver name : ov534
Card type : USB Camera-B4.04.27.1
Bus info : usb-musb-hdrc.1.auto-1.1
Driver version: 3.8.13
Capabilities : 0x85000001
Video Capture
Read/Write
Streaming
Device Capabilities
Device Caps : 0x05000001
Video Capture
Read/Write
Streaming
Priority: 2
Video input : 0 (ov534: ok)
Format Video Capture:
Width/Height : 640/480
Pixel Format : ‘YUYV’
Field : None
Bytes per Line: 1280
Size Image : 614400
Colorspace : SRGB
Streaming Parameters Video Capture:
Capabilities : timeperframe
Frames per second: 30.000 (30/1)
Read buffers : 2

User Controls

brightness (int) : min=0 max=255 step=1 default=0 value=0 flags=slider
contrast (int) : min=0 max=255 step=1 default=32 value=32 flags=slider
saturation (int) : min=0 max=255 step=1 default=64 value=64 flags=slider
hue (int) : min=-90 max=90 step=1 default=0 value=0 flags=slider
white_balance_automatic (bool) : default=1 value=1
exposure (int) : min=0 max=255 step=1 default=120 value=120 flags=inactive, volatile
gain_automatic (bool) : default=1 value=1 flags=update
gain (int) : min=0 max=63 step=1 default=20 value=20 flags=inactive, volatile
horizontal_flip (bool) : default=0 value=0
vertical_flip (bool) : default=0 value=0
power_line_frequency (menu) : min=0 max=1 default=0 value=0
sharpness (int) : min=0 max=63 step=1 default=0 value=0 flags=slider

Camera Controls

auto_exposure (menu) : min=0 max=1 default=0 value=0 flags=update

Sorry, forgot my comment please

Hi Derek.
firsly thank you very much. it was really useful for me.
but i have a doubt
if executing any of the programs, the answer is

cssp-camera 18000000.camera: dma_alloc_coherent of size 0 failed
VIDIOC_REQBUFS error 12, Cannot allocate memory

could you help me? THANK YOU

Hi Derek,
Thanks for your wonderful work and guidance.
I have a question here.
Could we put some texts or overlay some text on the video when v4l2 capture the real video? Like your code capture.c?
Thanks very much.
Yang

Hi Derek,

I followed your tutorials on grabbing picture frames from a webcam and also the one of cross-compiling, and they were indeed inspiring. I used my beaglebone to do the same, and it worked fine. But, I tried cross-compiling it on Eclipse IDE using arm-linux-gnueabi toolchain, it displayed errors on all the methods the contain “v4l2_”. For example: v4l2_open(), v4l2_mman(), v4l2_munmap(), v4l2_mmap(), v4l2_close() and v4l2_ioctl(). I later on changed the toolchain and used arm-angstrom-linux-gnueabi – hoping that it would solve the problem, since beaglebone is running angstrom linux. The IDE displayed errors that all the afore-mentioned methods were not defined. I would like to ask: were these errors due to the fact that these toolchains don’t have the library for these methods or the actual path for the library files were not properly used by the cross-compiler? I tried running it as a C++ program and also as a C program, I had the same errors.

I wrote other programs that did work fine. For example, the beaglebone is connected with a PIR sensor and a LED. When the sensor detects movement the LED goes on and if there is no movement after few seconds it goes off. This did work fine with the cross-compiler, and the program was written in C++.

I love Derek’s videos and have learned so much in so short a time. I’ve been battling a problem with grabber. Build went fine. Command line execution was easy. Conversion was easy. I can scp the files to my mac and look at them, awesome! So I went ahead and wrote a script that checks a value in a gpio folder to see if the PIR has been tripped and then execute grabber. Then it changes the file into a .jpg and transfers it to my mac. Works for hours, sometimes lots of hours, and then goes into a loop. I come home after work and find grabber on “top” in a tight loop. If I kill it, the script continues. Sometimes my device,

Cased:~# lsusb
Bus 001 Device 002: ID 1a40:0101 Terminus Technology Inc. 4-Port HUB
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 007: ID 046d:082d Logitech, Inc.
Bus 001 Device 004: ID 1a40:0101 Terminus Technology Inc. 4-Port HUB
Bus 001 Device 005: ID 0846:9030 NetGear, Inc. WNA1100 Wireless-N 150 [Atheros AR9271]

no longer shows up. My next step may be to look for an upgraded version of grabber, or maybe work capture into the home invasion detection software. Actually to date I have been merely taking pictures of my parrots.

If I come to Ireland, will you have a Guiness with me? It’s on my bucket list.