Difference between revisions of "Visual Bee Counter"
(→Software) |
|||
| Line 1: | Line 1: | ||
==[[Bee_Counter|Background]]== | ==[[Bee_Counter|Background]]== | ||
| + | An attempt to replicate an improve the techniques outlined in <i>Video Monitoring of Honey Bee Colonies at the Hive Entrance</i>, by Campbell, Mummert, and Sukthankar. | ||
| + | |||
| + | Optical and video bee counters. | ||
==Cameras== | ==Cameras== | ||
| + | Initial tests were conducted using these cameras: | ||
| + | |||
#Logitech 2500 | #Logitech 2500 | ||
#Logitech C310 | #Logitech C310 | ||
#Logitech HD920 | #Logitech HD920 | ||
| + | |||
| + | Need automatic light level control (auto iris). Don't really need auto focus. | ||
===Infrared (night time) Operation=== | ===Infrared (night time) Operation=== | ||
| + | |||
| + | ===Camera Enclosures=== | ||
==Software== | ==Software== | ||
| Line 13: | Line 22: | ||
##[http://docs.opencv.org/trunk/doc/py_tutorials/py_video/py_bg_subtraction/py_bg_subtraction.html Background Subtraction] | ##[http://docs.opencv.org/trunk/doc/py_tutorials/py_video/py_bg_subtraction/py_bg_subtraction.html Background Subtraction] | ||
#match an elliptical, graduated template at 16 orientations across each background-subtracted video frame. | #match an elliptical, graduated template at 16 orientations across each background-subtracted video frame. | ||
| − | |||
| − | |||
Revision as of 04:11, 25 December 2013
Contents
Background
An attempt to replicate an improve the techniques outlined in Video Monitoring of Honey Bee Colonies at the Hive Entrance, by Campbell, Mummert, and Sukthankar.
Optical and video bee counters.
Cameras
Initial tests were conducted using these cameras:
- Logitech 2500
- Logitech C310
- Logitech HD920
Need automatic light level control (auto iris). Don't really need auto focus.
Infrared (night time) Operation
Camera Enclosures
Software
- adaptive back-ground subtraction using a background model derived from a running average of the most recent 300 video frames.
- match an elliptical, graduated template at 16 orientations across each background-subtracted video frame.
