This paper deals with an image compression algorithm based on the three- dimensional Karhunen- Loeve transform (3D
KLT), whose task is to reduce time redundancy for an image data. There are many cases for which the reduction of time
redundancy is very efficient and brings perceptible bitstream reduction. This is very desirable for transfering image data
through telephone links, GSM networks etc. The time redundancy is very perceptible e.g. in camera security systems
where relative unchanging frames are very conventional. The time evolution of grabbed scene is reviews according an
energy content in eigenimages. These eigenimages are obtained in KLT for a suitable number of incoming frames.
Required number of transfered eigenimages and eigenvectors is determined on the basis of the energy content in
"3D KLT compression algorithm for camera security systems", Proc. SPIE 7138, Photonics, Devices, and Systems IV, 713820 (18 November 2008); doi: 10.1117/12.818065; https://doi.org/10.1117/12.818065