Multi-Scale Compressive Sensing Technology Revealed in InView’s Latest Patent

InView has invented a new more flexible approach to gathering compressive measurements that simultaneously collects data at multiple resolutions.

(Left) Hadamard matrix where the Kronecker product structure can be seen in the array of copies of matrix H12.(Right) The column-permuted Hadamard matrix.

(Left) Hadamard matrix where the Kronecker product structure can be seen in the array of copies of matrix H12.(Right) The column-permuted Hadamard matrix.

Sylvester-type Hadamard matrices are typically used as CS measurement patterns by reshaping rows of the matrix into 2D patterns that are displayed on a spatial light modulator and imprinted on the incoming scene in a CS camera.   InView takes advantage of the multiscale structure inherent in these matrices by using a permutation to group elements of the Hadamard matrix to specific locations on the SLM while introducing a random diagonal matrix to produce incoherence.

InView’s patent on techniques behind the creation of multi-scale measurement patterns is entitled, “Generating Modulation Patterns for the Acquisition of Multiscale Information in Received Signals.”  InView now has 15 issued patents and more applications in process. All of InView’s patents are described in our intellectual property white paper  InViewTechnologyCorp-Patents.

Our recent paper about other improved CS measurement techniques can be found at InViewCorp.com or at :  http://dx.doi.org/10.1117/12.2180474

Posted in Blog
Categories
Archives