System and Methods for Visual Sub-Band Decomposition of Image Signals


Tufts University investigators have developed novel methods to enhance, measure, fuse, and process images and videos based on Human Visual System Characteristics (HVSC). Human Visual System (HVS) based image decomposition emulates the way in which human eyes respond to visual stimulus.



Digital images are subject to a wide variety of distortions during acquisition, processing, compression, storage, transmission and reproduction, any of which may result in a degradation of visual quality. In practice, subjective evaluation is usually too inconvenient,  time-consuming and expensive. The goal of image quality assessment is to supply quality metrics that can predict perceived image quality automatically by using computational devices. This opportunity has applications in the multi-billion dollar multimedia analytics market.



This new approach addresses challenges of current enhancement solutions including difficult cases of varying scene illumination and face recognition. It can process the images in real time and can operate in a wire and/or wireless transmission environment or embedded in a hardware, firmware, or software which is embedded in another system or application.



A base image is used as an input and generates sub band decomposed images that may include variable level sub-band decomposition. The resulting sequence of sets of sub images may be processed independently and may be transmitted, analyzing and displaying separately. The sub-band decomposed images can be stored and retrieved to be applied to any number of different processing cycles and incorporate with other decomposition methods. The decomposed images may then be fused back together to reconstruct an output image. These resulting sets of sub images can be used/incorporated as input to a variety of image processing techniques such as, enhancement, edge detection, filtering, compression, recognition, denosing, restoration, digital forensic, security, and information hiding.



-object detection

-object recognition,

-security, medical and digital image enhancement



-Automated image processing

-Superior image enhancement – detail preservation

-Improved face recognition

-Improved enhancement at varying illuminations


IP STATUS  US Patent 8,285,076, Issued October 9, 2012


Licensing Contact

Frannie Raede