You are here

Category
Technology Name
Briefcase
Scientist
1522
A method for enhancing the spatial and or temporal resolution (if applicable) of an input signal such as images and videos.   Many imaging devices produce signals of unsatisfactory resolution (e.g. a photo from a cell-phone camera may have low spatial resolution or a video from a web camera may have...

A method for enhancing the spatial and or temporal resolution (if applicable) of an input signal such as images and videos.

 

Many imaging devices produce signals of unsatisfactory resolution (e.g. a photo from a cell-phone camera may have low spatial resolution or a video from a web camera may have both spatial and temporal low resolution). This method applies digital processing to reconstruct more satisfactory high resolution signals.

 

Previous methods for Super-Resolution (SR) require multiple images of the same scene, or else an external database of examples. This method provides the ability to perform SR from a single image (or a single visual source). The algorithm exploits the inherent local data redundancy within visual signals (redundancy both within the same scale, and across different scales).

 

Examples of the methods' capabilities can be found here: http://www.wisdom.weizmann.ac.il/~vision/SingleImageSR.html

 

Applications


  • Enhancing the spatial resolution of images

  • Enhancing the spatial and or temporal resolution of video sequences

  • Enhancing the spatial and or temporal resolution (if applicable) of other signals (e.g., MRI, fMRI, ultrasound, possibly also audio, etc.)

 


Advantages


  • No need for multiple low resolution sources or the use of an external database of examples.

  • Superior results are produced due to exploitation of inherent information in the source signal.


Technology's Essence


The framework combines the power of classical multi image super resolution and example based super resolution. This combined framework can be applied to obtain super resolution from as little as a single low-resolution signal, without any additional external information. The approach is based on an observation that patches in a single natural signal tend to redundantly recur many times inside the signal, both within the same scale, as well as across different scales.

Recurrence of patches within the same scale (at subpixel misalignments) forms the basis for applying the 'classical super resolution' constraints to information from a single signal. Recurrence of patches across different (coarser) scales implicitly provides examples of low-resolution / high-resolution pairs of patches, thus giving rise to 'example-based super-resolution' from a single signal (but without any external database or any prior examples).

+
  • Prof. Michal Irani
1121
A method for aligning video images according to sequence. The problem of image alignment has been extensively studied, and successful approaches have been developed for solving this problem. However, these approaches turn out as problematic when there is insufficient overlap between the two images to...

A method for aligning video images according to sequence. The problem of image alignment has been extensively studied, and successful approaches have been developed for solving this problem. However, these approaches turn out as problematic when there is insufficient overlap between the two images to allow extraction of common image properties, i.e., when there is no sufficient similarity (e.g., gray-level, frequencies, statistical) between the two images. Whereas two individual images cannot be aligned when there is no spatial overlap between them, this is not the case when dealing with image sequences. The outlined technology consists of fusion and alignment of discrete, non-overlapping moving images from different sources, by aligning spatio-temporal changes in each sequence rather than in each image.

Applications


  • Multi-sensor image alignment for multi-sensor fusion
  • Alignment of images (sequences) obtained at significantly different zooms (can be useful in surveillance applications)
  • Generation of wide-screen movies from multiple non-overlapping narrow field-of-view movies (such as in IMAX movies) 
  • Alignment and integration of information across video sequences to exceed the physical visual limitations of any individual sensor (e.g., dynamic range, spectral range, spatial resolution, temporal resolution, etc). ~

Advantages


  • Useful for spatially non-overlapping sequences
  • Useful in cases which are inherently difficult for standard image alignment techniques, such as when there is insufficient common spatial information across the two sequences

Technology's Essence


An image sequence contains much more information than any individual image frame does. In particular, temporal changes in a video sequence (e.g., due to camera motion) do not appear in any individual image frame, but are encoded between video frames. When these temporal changes are common to the two sequences, then these sequences can be aligned both in time and in space, even if there is no common spatial information whatsoever. The need for coherent visual appearance, which is a fundamental assumption in image alignment methods, is replaced in this invention with the requirement of coherent temporal behavior. This can be achieved by attaching the two video cameras closely to each other (so that their centers of projections are very close), and moving them jointly in space (e.g., such as when the two cameras are mounted on a moving platform or rig).

 

Click here for additional information
Click here to visit Prof. Irani`s Homepage

+
  • Prof. Michal Irani
1381

Applications


The new method for detecting irregularities has many applications which include:

  1. Detecting suspicious and/or salient behaviors in video
  2. Attention and saliency in images
  3. Detecting irregular tissue in medical images
  4. Automatic visual inspection for quality assurance (e.g., detecting defects in goods)
  5. Generating a video summary/synopsis
  6. Intelligent fast forward
  7. Non-visual data

    Technology's Essence


    Researchers at the Weizmann Institute have developed a new method for detecting irregularities based only on few regular examples, without any assumed models. In the new method the validity of data is determined as a process of constructing a puzzle: one tries to compose a new observed image region or a new video segment (''the query'') using chunks of data (''pieces of puzzle'') extracted from previous visual examples (''the database''). Regions in the observed data which can be composed using large contiguous chunks of data from the database are considered very likely, whereas regions in the observed data which cannot be composed from the database (or can be composed, but only using small fragmented pieces) are regarded as unlikely/suspicious. The problem is posed as an inference process in a probabilistic graphical model. The invention also includes an efficient algorithm for detecting irregularities. Moreover, the same method can also be used for detecting irregularities/anomalies within data without any prior examples, by learning the notion of regularity/irregularity directly from the query data itself.

    Click here to see additional features

+
  • Prof. Michal Irani

Pages