IP Library Granted Patent US 8,836,809
Granted Patent B2
US 8,836,809 · App. 13/620,963 · Granted Sep 16, 2014

Quad-core image processor for facial detection

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 8,836,809
App. No.
13/620,963
Granted
Sep 16, 2014
Kind
B2
Abstract

A quad-core processor for a hand held device with a CMOS image sensor to capture a scene. The quad-core processor has an image sensor interface for receiving data from the CMOS image sensor and four processing units for simultaneously processing the data. The image sensor and the four processing units being incorporated onto a single chip and the processing units are configured to detect faces within the scene.

Claims (40)

1. A processor for a hand-held device, the processor comprising:

an image sensor interface for receiving data from an image sensor configured to capture a scene; and,

a plurality of processing units for simultaneously processing the data, the image sensor interface and the plurality of processing units being incorporated onto a single chip;

wherein, the plurality of processing units are configured to:

detect one or more faces within an image corresponding to the data, and

apply a warp to the one or more faces to manipulate the image using a warp map, wherein the warp map comprises an array of values that set out coordinates of an input image that map to corresponding coordinates in an output image, wherein the input image and the output image have different image dimensions.

2. The processor according to claim 1 , wherein the plurality of processing units is configured to determine hues and to perform local feature analysis of the data to detect the one or more faces.

3. The processor according to clam 2 , further comprising:

an input FIFO (first in, first out) for receiving the data from the image sensor interface and inputting the data to the plurality of processing units, and

an output FIFO for receiving processed data from the plurality of processing units.

4. The processor according to clam 3 , further comprising a central processing unit and a data bus connecting the central processing unit to the input FIFO and the output FIFO, the central processing unit providing a processing core to operatively control the plurality of processing units.

5. The processor according to claim 4 , further comprising a data cache to store the processed data until required by the central processing unit, wherein the data cache is integrated onto the single chip with the plurality of processing units.

6. The processor according to claim 5 , further comprising:

a display interface communicatively coupled to a display device, wherein the display interface is integrated onto the single chip with the plurality of processing units and connected to the data cache for reading the processed data and transmitting the processed data to the display device.

7. The processor according to claim 6 , wherein each of the plurality of processing units has at least one address generator for the processed data transmitted to the data cache.

8. The processor according to claim 4 , further comprising:

a USB (universal serial bus) interface integrated onto the single chip with the plurality of processing units, wherein the USB interface is communicatively coupled to a USB port for communication with an external device, such that the USB interface connects to the central processing unit via the data bus for communication with the external device.

9. The processor according to clam 1 , further comprising a crossbar switch for interconnecting each of the plurality of processing units.

10. The processor according to claim 1 , wherein the plurality of processing units perform one or more processing tasks selected from:

rotating the image corresponding to the data;

color converting the image corresponding to the data; and

dithering the image corresponding to the data.

11. The processor according to claim 1 , wherein applying the warp comprises:

scaling the warp map to correspond to dimensions of the output image;

determining a region of input image pixels corresponding to each output image pixel; and

calculating each output image pixel based on the scaled warp map and the corresponding region of input image pixels.

12. The processor according to claim 1 , wherein the warp map has dimensions that are different from dimensions of the input image and dimensions of the output image.

13. An apparatus comprising:

a network interface;

an image capture device;

a display; and

a plurality of interconnected processing units arranged to execute one or more programs for operation of the image capture device, the display, and the network interface,

wherein the one or more programs executed by the processing units provide effects including detecting one or more faces within an image captured by the image capture device and applying a warp to the one or more faces to manipulate the image using a warp map, wherein the warp map comprises an array of values that set out coordinates of an input image that map to corresponding coordinates in an output image, wherein the input image and the output image have different image dimensions.

14. The apparatus according to claim 13 , wherein the image capture device comprises a CMOS image sensor.

15. The apparatus according to claim 13 , wherein the processing units are configured to determine hues and to perform local feature analysis of the image to detect the one or more faces.

16. The apparatus according to claim 13 , wherein applying the warp comprises:

scaling the warp map to correspond to dimensions of the output image;

determining a region of input image pixels corresponding to each output image pixel; and

calculating each output image pixel based on the scaled warp map and the corresponding region of input image pixels.

17. The apparatus according to claim 13 , wherein the warp map has dimensions that are different from dimensions of the input image and dimensions of the output image.

Assignments (3)
CHANGE OF NAME Recorded Oct 5, 2017
From: GOOGLE INC.
To: GOOGLE LLC
Reel/Frame 044129/0001 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 14, 2014
From: SILVERBROOK, KIA
To: SILVERBROOK RESEARCH PTY LTD
Reel/Frame 031967/0855 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 18, 2013
From: SILVERBROOK RESEARCH PTY LTD
To: GOOGLE INC.
Reel/Frame 029824/0191 →