Detection, recognition and tracking of small flying targets

Recognition of hidden pattern with background (2007)


This target masking method incorporates our recently introduced and patented relative focus map estimation method for automatically separating less- and more- focused regions of images or video frames, for both smaller and larger regions. Then, a pixel intensity history graphs and time-based region histograms are used for moving object separation, mainly for small moving objects. Overall, the method is new in using a combination of the above two approaches with a Stauffer-Grimson background suppressing and region tracking method in order to create a method which can produce target masks on low resolution noisy input from small to large object sizes.


Focus region extraction.



Stauffer-Grimson-based masking.



Local histogram-statistics based tracking.Each element of the matrix is a diagram with horizontal axis being the time, vertical axis being the normalized intensity value of the respective pixel. Changing graph values show the passing through of a moving object.



Masks obtain from histogram statistics.



Masks obtained with the combined focus-local histogram statistics steps.


Extraction, categorization and unusual motion signaling of small moving objects (2009)

The paper presents an automatic approach for small moving object detection, categorization and unusual motion pattern signaling on camera feeds on sky background. The method uses a local blind deconvolution based foreground detector for small object mask and contour edge extraction, spatio-temporal localized histogram evaluation for object classification, and a hidden Markov model based evaluation for learning usual motions and signaling unusual motion patterns. The method is able to mask moving objects, fit them into learned categories and signal unexpected motion behavior.