[Bug 629244] [opencv] Add motion detection element

GStreamer (bugzilla.gnome.org) bugzilla at gnome.org
Sun Jul 17 09:47:11 PDT 2011


https://bugzilla.gnome.org/show_bug.cgi?id=629244
  GStreamer | gst-plugins-bad | git

Nicola <lists> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #189667|0                           |1
        is obsolete|                            |
 Attachment #189668|0                           |1
        is obsolete|                            |
 Attachment #189698|0                           |1
        is obsolete|                            |

--- Comment #131 from Nicola <lists at svrinformatica.it> 2011-07-17 16:47:08 UTC ---
Created an attachment (id=192132)
 View: https://bugzilla.gnome.org/attachment.cgi?id=192132
 Review: https://bugzilla.gnome.org/review?bug=629244&attachment=192132

use g_mutex and not pthread mutex, solved a lock bug

Hi,

I tested the plugin again and it seems to work fine, there are no compile
warnings anymore, valgrind show two possibly lost

==4129== 168 bytes in 1 blocks are possibly lost in loss record 2,177 of 2,304
==4129==    at 0x4C28FAC: malloc (vg_replace_malloc.c:236)
==4129==    by 0xBBD0B24: cv::fastMalloc(unsigned long) (in
/usr/lib/libcxcore.so.2.1.0)
==4129==    by 0xBC16E21: cvCloneImage (in /usr/lib/libcxcore.so.2.1.0)
==4129==    by 0xAC406C7: MotionCells::performDetectionMotionCells(_IplImage*,
double, double, int, int, long, bool, bool, int, motionmaskcoordrect*, int,
motioncellidx*, cellscolor*, int, motioncellidx*, long, char*, bool, int)
(MotionCells.cpp:218)
==4129==    by 0xAC3D5D9: perform_detection_motion_cells
(motioncells_wrapper.cpp:44)
==4129==    by 0xAC3C698: ??? (gstmotioncells.c:979)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B613C5: gst_base_transform_chain (gstbasetransform.c:2458)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B613C5: gst_base_transform_chain (gstbasetransform.c:2458)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B59697: gst_base_src_loop (gstbasesrc.c:2516)
==4129==    by 0x4EB2BEF: gst_task_func (gsttask.c:318)
==4129==    by 0x55D7B15: ??? (in
/lib/x86_64-linux-gnu/libglib-2.0.so.0.2800.6)
==4129==    by 0x55D53E3: ??? (in
/lib/x86_64-linux-gnu/libglib-2.0.so.0.2800.6)
==4129==    by 0x5860D8B: start_thread (pthread_create.c:304)
==4129==    by 0x5B5E04C: clone (clone.S:112)
==4129== 
==4129== 921,624 bytes in 1 blocks are possibly lost in loss record 2,304 of
2,304
==4129==    at 0x4C28FAC: malloc (vg_replace_malloc.c:236)
==4129==    by 0xBBD0B24: cv::fastMalloc(unsigned long) (in
/usr/lib/libcxcore.so.2.1.0)
==4129==    by 0xBC04447: cvCreateData (in /usr/lib/libcxcore.so.2.1.0)
==4129==    by 0xBC16E92: cvCloneImage (in /usr/lib/libcxcore.so.2.1.0)
==4129==    by 0xAC406C7: MotionCells::performDetectionMotionCells(_IplImage*,
double, double, int, int, long, bool, bool, int, motionmaskcoordrect*, int,
motioncellidx*, cellscolor*, int, motioncellidx*, long, char*, bool, int)
(MotionCells.cpp:218)
==4129==    by 0xAC3D5D9: perform_detection_motion_cells
(motioncells_wrapper.cpp:44)
==4129==    by 0xAC3C698: ??? (gstmotioncells.c:979)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B613C5: gst_base_transform_chain (gstbasetransform.c:2458)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B613C5: gst_base_transform_chain (gstbasetransform.c:2458)
==4129==    by 0x4E8C53B: gst_pad_push (gstpad.c:4684)
==4129==    by 0x8B59697: gst_base_src_loop (gstbasesrc.c:2516)
==4129==    by 0x4EB2BEF: gst_task_func (gsttask.c:318)
==4129==    by 0x55D7B15: ??? (in
/lib/x86_64-linux-gnu/libglib-2.0.so.0.2800.6)
==4129==    by 0x55D53E3: ??? (in
/lib/x86_64-linux-gnu/libglib-2.0.so.0.2800.6)
==4129==    by 0x5860D8B: start_thread (pthread_create.c:304)
==4129==    by 0x5B5E04C: clone (clone.S:112)

however I monitored the ram usage in /proc/<pid>/status for about an hour and
the memory usage is constant, maybe we need some opencv suppression file.

I think if the latest property I added for the Roman use case works fine this
plugin is "bad" enough for the inclusion. Stefan, Tim could you please review
the plugin and tell us what need to be done to have motioncells in -bad?

thanks
Nicola

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.


More information about the gstreamer-bugs mailing list