Foveated compressive sensing system

• US 9,230,302 B1
• Filed: 03/11/2014
• Issued: 01/05/2016
• Est. Priority Date: 03/13/2013
• Status: Active Grant

• Current Assignee: HRL Laboratories LLC (The Boeing Co.)
• Original Assignee: HRL Laboratories LLC (The Boeing Co.)
• Inventors: Owechko, Yuri, Ni, Kang-Yu, Rao, Shankar R.
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US14/204,028
• Time in Patent Office: 665 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06T 1/0007 Image acquisition
• CPC Class Codes: G06T 3/04 Context-preserving transformations, e.g. by using an importance map panospheric to cylindrical image transformations G06T3/12
• CPC Class Codes: G06V 10/22 by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
• CPC Class Codes: G06V 10/25 Determination of region of interest [ROI] or a volume of interest [VOI]
• CPC Class Codes: H03M 7/3062 Compressive sampling or sensing

Abstract

Described is a system for foveated compressive sensing. The system is configured to receive an input image f of a scene and initialize a measurement matrix. Global measurements are then performed, with a lower resolution image of the scene thereafter reconstructed. Task salient regions are extracted from the low resolution image. Thereafter, the system estimates a task-specific operator and detects regions-of-interest (ROI) based on the task salient regions. An ROI-adapted and foveated measurement matrix is then generated. Local measurements are then performed on task-relevant ROIs. A higher resolution image can then be reconstructed of the scene to allow for identification of objects in the ROI.

Claims

1. A system for foveated compressive sensing, the system comprising:

one or more processors and a memory, the memory being a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of;

receiving an input image f of a scene;

initializing a measurement matrix;

performing global measurements of the input image f;

reconstructing a lower resolution image of the scene;

extracting task salient regions from the low resolution image;

estimating a task-specific operator and detecting regions-of-interest (ROI) based on the task salient regions;

generating an ROI-adapted and foveated measurement matrix;

performing local measurements on task-relevant ROIs; and

reconstructing a higher resolution image of the scene to allow for identification of objects in the ROI.

2. The system as set forth in claim 1, wherein initializing the measurement matrix further includes operations of:

defining a first task;

implementing a task prior; and

initializing the measurement matrix.

3. The system as set forth in claim 2, wherein in initializing a measurement matrix, the input image f is focused on a spatial light modulator (SLM) that processes image patches in parallel and generates rows of a foveated measurement matrix AP_f serially.

4. The system as set forth in claim 3, wherein at each time step j, a product of the input image f and row j of AP_f, is focused on a detector array having a plurality of detectors, such that each detector in the detector array performs a separate patch measurement by spatially integrating AP_ff in its local patch area.

5. A computer program product for foveated compressive sensing, the computer program product comprising:

a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of;

receiving an input image f of a scene;

initializing a measurement matrix;

performing global measurements of the input image f,

reconstructing a lower resolution image of the scene;

extracting task salient regions from the low resolution image;

estimating a task-specific operator and detecting regions-of-interest (ROI) based on the task salient regions;

generating an ROI-adapted and foveated measurement matrix;

performing local measurements on task-relevant ROIs; and

reconstructing a higher resolution image of the scene to allow for identification of objects in the ROI.

6. The computer program product as set forth in claim 5, wherein initializing the measurement matrix further includes operations of:

defining a first task;

implementing a task prior, and

initializing the measurement matrix.

7. The computer program product as set forth in claim 6, wherein in initializing a measurement matrix, the input image f is focused on a spatial light modulator (SLM) that processes image patches in parallel and generates rows of a foveated measurement matrix AP_f serially.

8. The computer program product as set forth in claim 7, wherein at each time step j, a product of the input image f and row j of AP_f, is focused on a detector array having a plurality of detectors, such that each detector in the detector array performs a separate patch measurement by spatially integrating AP_ff in its local patch area.

9. A computer implemented method for foveated compressive sensing, the method comprising an act of:

causing one or more processors to execute instructions encoded on a non-transitory computer-readable medium, such that upon execution, the one or more processors perform operations of;

receiving an input image f of a scene;

initializing a measurement matrix;

performing global measurements of the input image f,

reconstructing a lower resolution image of the scene;

extracting task salient regions from the low resolution image;

estimating a task-specific operator and detecting regions-of-interest (ROI) based on the task salient regions;

generating an ROI-adapted and foveated measurement matrix;

performing local measurements on task-relevant ROIs; and

reconstructing a higher resolution image of the scene to allow for identification of objects in the ROI.

10. The method as set forth in claim 9, wherein initializing the measurement matrix further includes operations of:

defining a first task;

implementing a task prior; and

initializing the measurement matrix.

11. The method as set forth in claim 10, wherein in initializing a measurement matrix, the input image f is focused on a spatial light modulator (SLM) that processes image patches in parallel and generates rows of a foveated measurement matrix AP_f serially.

12. The method as set forth in claim 11, wherein at each time step j, a product of the input image f and row j of AP_f, is focused on a detector array having a plurality of detectors, such that each detector in the detector array performs a separate patch measurement by spatially integrating AP_ff in its local patch area.