Performing a histogram using an array of addressable registers

• US 9,122,954 B2
• Filed: 10/01/2013
• Issued: 09/01/2015
• Est. Priority Date: 10/01/2013
• Status: Active Grant

• Current Assignee: MobilEye Vision Technologies Ltd. (Intel Corporation)
• Original Assignee: MobilEye Vision Technologies Ltd. (Intel Corporation)
• Inventors: Srebnik, Daniel, Dogon, Gil
• Primary Examiner(s): COUSO, JOSE L
• Application Number:
  US14/043,030
• Publication Number:
  US 20150093028A1
• Time in Patent Office: 700 Days
• US Class Current: 1/1
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining

Abstract

Parallel processing of an image using an array of addressable registers. Image features are extracted from the image. The image features are storable as data. According to respective values of a sorting key derived from a parameter of the data, the image features are sorted into N buckets. Using an array of M addressable registers, where M is less than N, the data are summed within the buckets to perform a histogram of the image features, according to values of a histogram key derived from said a parameter of the data.

Claims

1. A method for parallel processing of an image using processing circuitry including an array of addressable registers, the method comprising:

extracting a plurality of image features from the image, wherein the image features are storable as data;

according to respective values of a sorting key, wherein said sorting key is derived from at least one parameter of the data, sorting the image features into a plurality of N buckets; and

using an array of M addressable registers, where M is less than N, according to values of a histogram key derived from said at least one parameter of the data, summing said data within said buckets thereby performing a histogram of the image features.

2. The method of claim 1, wherein the respective data of the image features are storable in binary storage in data words of I bits, the method, further comprising:

said sorting the image features into N=2^J buckets, wherein J is an integer greater than one and less than I; and

using the array of at least M=2^K addressable registers, wherein K is an integer greater or equal to the difference I−

J between integer I and integer J, said summing said data within said buckets thereby said performing said histogram of the image features.

3. The method of claim 1, wherein the addressable registers are configured to manipulate data stored in a read-modify-write sequence.

4. The method of claim 1, wherein said sorting includes a sorting by radix, further comprising:

determining the radix over which to sort by performing a histogram of the data over the radix, thereby preliminarily determining a resolution of said sorting key.

5. The method of claim 1, wherein said data of said image features are selected from the group consisting of:

a color coordinate, at least one of a direction and magnitude of a gradient of a color coordinate, a grayscale value, at least one of a direction and magnitude of a gradient of a grayscale value.

6. The method of claim 1, wherein said data of said image features include data of edge features from a plurality of picture elements of the image selected from the group consisting of:

a horizontal position coordinate with respect to an origin in the image, a vertical position coordinate with respect to said origin, intensity, horizontal intensity gradient and vertical intensity gradient.

7. The method of claim 6, further comprising:

computing for at least one of said picture elements a distance between said at least one picture element and said origin based on said edge features; and

computing for said picture elements a direction angle relative to a coordinate axis in said image;

wherein said at least one parameter of the data is selected from the group consisting of;

said distance, said direction angle, a function of said distance and a function of said direction angle.

8. The method of claim 7, wherein said sorting key includes said direction angle used for said sorting into said buckets and said histogram key includes said distance used for said summing said data associated with like distances within said buckets.

9. The method of claim 7, wherein said sorting key includes said distance used for said sorting by distance into said buckets and said histogram key includes said direction angle used for said summing said data associated with like direction angles within said buckets.

10. The processing circuitry including the array of addressable registers configured to perform the method of claim 1, the method comprising:

extracting a plurality of image features from the image, wherein the image features are storable as data;

according to respective values of a sorting key, wherein said sorting key is derived from at least one parameter of the data, sorting the image features into a plurality of N buckets; and

using an array of M addressable registers, where M is less than N, according to values of a histogram key derived from said at least one parameter of the data, summing said data within said buckets thereby performing a histogram of the image features.