Lossless Codec for Infrared or Monochrome Video
Infrared cameras such as the Electophysics PV320 provide low cost, high definition infrared images. Infrared cameras can be used to detect disturbed earth, spot heating/cooling system inefficiencies, and track targets at night. Due to the physical nature of the data collected by these cameras, it would be very desirable to be able to store the data without distortion or errors. An uncompressed frame of video is 153,600 bytes and at 30 frames per second, it would take less than six hours to completely fill a 100GB hard disk if the frame data was written in its raw form. A codec which provides lossless data compression is needed in this case for efficient data collection and storage.
Researchers at the Johns Hopkins Applied Physics Laboratory (JHU/APL) have developed a lossless (i.e. the original 14-bit data can be exactly resorted from the recorded video stream) codec to compress this type of video data. A capture program has been written using this codec to provide data compression. The codec uses a technique called arithmetic coding to compress video data. Arithmetic coding represents a stream of compressed data as a single floating-point number of arbitrary precision. Because real floating-point numbers have very specific, limited precision, the algorithm is actually implemented using integer arithmetic. This codec can compress this data at a 50% ratio, which is comparable to general purpose codecs that are commonly available but are lossy. Computational loading for this codec is low enough to permit real-time recording for the intended infrared sensor.
We are seeking a commercial transition licensee, either a start-up or an existing company, focusing on non-US Government customers who can successfully commercialize this IP.
Patent Status: U.S. patent(s) 8,817,875 issued.
Mr. E. Chalfin
Phone: (443) 778-7473