Bee counter software

From HiveTool
Revision as of 03:48, 21 March 2015 by Paul (talk | contribs) (Summary)
Jump to: navigation, search
Threshold image.
Filtered (smothed) threshold image.
Show contours
Show contours and tracks.
Same but input image is off
White track has terminated in green area and will count as out


beeTrack1 is a c++ program using openCV libraries. Here is an example of it processing video:

At frame 500, the contours are turned off, showing only the tracks. At 700, the input image (background) is turned off, showing only the contours and tracks.

When a track (series of brown dots for a single bee) is counted, it will turn green (out) or red (in) as it is counted, just before it is erased.

Processing steps for each frame:

  1. Convert image to grey scale.
  2. Threshold based on grey value. Result is binary image where dark areas are 0s and light areas are 1s.
  3. Invert threshold image so bees that were black (0) are now white (1).
  4. Filter to smooth the noise. (The exact process changes from version to version but currently:
    1. Erode the threshold image twice to remove legs, etc and try to separate the bees.
    2. Blur the image to smoth it.
  5. Draw a color contour around each object and color code it based on size (area).
  6. Check each existing track to see if one falls within a contour of an object.
  7. If so, append the center coordinates of the object to the track. If not, start a new track.
  8. When the track ends, increment the IN or OUT counter depending on where the track starts, ends and length.

Note that these contour and track colors are based on the area of the contours. The number of bees is estimated by dividing the contour area by an arbitrary bee size. The actual number of bees depends on how tight the group of bee is and settings that affect the size of the contours.

1 beeBrown
2 beesOrange
3 beesYellow
4 beesBlue
5 beesViolet


Open Computer Vision (openCV) is an open source computer vision and machine learning software library. The library has more than 2500 optimized algorithms, which includes a comprehensive set of both classic and state-of-the-art computer vision and machine learning algorithms.

Kyle Hounslow has posted several YouTube videos on using openCV for motion tracking:

Note that what most videos call tracking, is really just object identification. Tracking the object from frame to frame is trivial if there is only one object, but becomes harder and more error prone if there are multiple objects, especially if they can touch or occlude each other.