Multi-Function Summing Machine

• US 20110280495A1
• Filed: 05/14/2010
• Published: 11/17/2011
• Est. Priority Date: 05/14/2010
• Status: Active Grant

• Current Assignee: MobilEye Vision Technologies Ltd. (Intel Corporation)
• Original Assignee: Mobileye Technologies Limited (Intel Corporation)
• Inventors: Sixsou, Emmanuel, Navon, Mois
• Granted Patent: US 8,538,205 B2
• Time in Patent Office: Days
• US Class Current: 382/276
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining

Abstract

A system for processing an image including multiple pixels and intensity data thereof. An image memory is adapted for storing the image. An arithmetic core is connectible to the image memory and adapted for inputting the intensity data. The arithmetic core includes a multiple function processing units. One or more of the function processing units includes (i) a processing core adapted for computation of a function of the intensity data and for producing results of the computation, (ii) a first and (iii) a second accumulator for summing the results; and storage adapted to store the results. The function processing units are configured to compute the functions in parallel and sum the results simultaneously for each of the pixels in a single clock cycle.

Claims

1. A system for processing an image frame including a plurality of pixels and intensity data thereof, the system comprising:

an image memory adapted for storing the image frame; and

an arithmetic core connectible to said image memory and adapted for inputting the intensity data, said arithmetic core including a plurality of function processing units,

wherein said function processing units include

a processing core adapted for computation of a function of the intensity data and for producing a result of said computation; and

an accumulator for summing said result.

2. The system, according to claim 1, wherein said function processing units are configured to compute said functions in parallel and sum said results simultaneously for each of the pixels in a single clock cycle.

3. The system according to claim 1, further comprising

storage operatively attached to said arithmetic core and adapted to store said result.

4. The system according to claim 1, wherein said accumulator is a first accumulator, wherein said function processing units further include a second accumulator, and wherein said accumulator and said second accumulator are connected so that said result of said accumulator is input into said second accumulator.

5. The system according to claim 4, wherein an output of said second accumulator becomes a final result.

6. The system according to claim 1, wherein said processing core is configured to perform said computation of said function in a window of the image frame.

7. The system according to claim 1, wherein said function processing units are configured to compute said functions in parallel.

8. The system according to claim 7, further comprising:

a control module operatively connectible to said arithmetic core and adapted to specify said functions to be processed respectively in different windows of the image frame.

9. The system according to claim 8, wherein said control module is adapted to specify said different windows of the image frame.

10. The system according to claim 8, wherein said control module has an operation mode bit adapted to select window size and wherein said different windows are of different size.

11. The system according to claim 8, wherein said control module is adapted to specify overlap between said different windows.

12. A method for processing an image frame including a plurality of pixels and intensity data thereof, in a system including an image memory storing the image frame, the method comprising:

specifying an image window in the image frame; and

during a single read of the intensity data of the image frame, summing a plurality of functions of the intensity data in the image window by calculating said functions while incrementing over the pixels of the image window,

wherein said calculating of said functions is performed in parallel for each of the pixels in a single clock cycle and said summing of said functions is performed simultaneously.

13. The method according to claim 12, wherein, for each of said functions, said summing is performed by:

(i) first accumulating in a first accumulator a result of said calculating,

(ii) upon reaching an edge of the image window inputting said result into a second accumulator.

14. The method according to claim 13, wherein said summing further comprises:

(iii) continuing to increment over said image window until all the pixels of the image window are read; and

(iv) a second calculating and writing said result as a final result for the image window.

15. The method according to claim 14, wherein the final results for said functions are available at the same time to a plurality of driver assistance systems.

16. The method according to claim 12, wherein said image window includes a plurality of image windows, wherein said image windows are of different size.

17. The method according to claim 12, wherein said image window includes a plurality of image windows, wherein said image windows overlap.

18. A system for processing an image, the system adapted for performing the method according to claim 12, the system comprising:

the image memory adapted for storing the image; and

an arithmetic core connectible to the image memory and adapted for inputting the intensity data, said arithmetic core including a plurality of function processing units each including;

a processing core adapted for respective computations of the functions of the intensity data and for producing results of said computations, and an accumulator for summing the results of said computations.