Image signal analyzing system

• US 5,500,907 A
• Filed: 06/24/1992
• Issued: 03/19/1996
• Est. Priority Date: 06/25/1991
• Status: Expired due to Term

• Current Assignee: Fuji Xerox Company Limited (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Company Limited (Fujifilm Holdings Corporation)
• Inventors: Suzuki, Kazuhiro, Koshi, Yutaka, Kunitake, Setsu, Kamizawa, Koh
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): KELLEY, CHRISTOPHER S
• Application Number:
  US07/903,284
• Time in Patent Office: 1,364 Days
• Field of Search: 382/16, 382/21, 382/37, 358/136, 348/417, 348/409
• US Class Current: 382/240
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: H04N 19/94 Vector quantisation

Abstract

An image signal analyzing system capable of accomplishing shape analysis for processing or coding an image with easy, by space division using two orthogonal representative vectors, or by differential representative vectors given by the difference between representative vectors. With such a system, inner products between mean separated blocks produced by a mean separated unit and a differential representative vector set stored in a memory in advance are computed by inner product computing unit disposed into a binary tree shape having n stages. Positive/negative signs expressing resulting inner products in the respective stages of the binary tree computed by the inner product computing unit are held, as indexes, by n index holding means sequentially, and an index representing the shape of input image information is obtained from the n indexes held by the index holding unit.

Claims

1. An image signal analyzing system comprising:

a block extracting means for sequentially extracting blocks including a predetermined number of pixels from an input image;

a storage means for storing a set of differential representative vectors, each obtained from a difference between representative vectors at a same stage of a binary tree having a plurality of stages of said representative vectors;

a plurality of stages of inner product computing means for performing computation of inner products between said extracted blocks and said respective differential representative vectors of said plurality of stages; and

an inner product result holding means for holding results of said inner products obtained for every stage of said binary tree by said inner product computing means, as an index expressing a shape of said input image.

2. The image signal analyzing system of claim 1 wherein a set of representative vectors obtained by establishing representative vectors of the first and second stages so as to show two desired orthogonal directions expressing a spatial direction of said input image to thereby produce representative vectors of the succeeding stages, are used as the set of representative vectors to be stored in the form of a difference in every stage of said binary tree in said storage means.

3. The image signal analyzing system of claim 2 wherein said two orthogonal directions are vertical and horizontal directions, respectively.

4. The image signal analyzing system of claim 2 wherein said two orthogonal directions are left-inclined and right-inclined directions, respectively.

5. The image signal analyzing system of claim 2 wherein said two orthogonal directions are vertical and horizontal directions of second-order harmonic respectively, in which no values of components in vectors increase monotonously but increase and decrease twice.

6. The image signal analyzing system of claim 1 wherein representative vectors produced to establish a symmetrical relationship between two representative vectors in each stage are used, only one of the two representative vectors in said symmetrical relationship in each stage being stored in the form of a difference, in said storage means.

7. The image signal analyzing system of claim 1 further comprising a selection means for selectively reading from said storage means a differential representative vector to thereby obtain an operator of an inner product of said inner product computing means in a certain stage in accordance with an inner product result held by said inner product result holding means in the next stage.

8. The image signal analyzing system of claim 7 further comprising a generation and input means for generating a differential representative vector to be supplied to an input of said inner product computing means in said next stage on the basis of a differential representative vector on one side of said symmetrical relationship read selectively from said storage means by said selection means, and for supplying the thus generated differential representative vector into the input of said inner product computing means in the next stage.

9. The image signal analyzing system of claim 8 wherein said generation and input means for supplying said inner product computing means in the next stage is arranged so that:

in the case where said differential representative vector on one side of said symmetrical relationship read from said storage means is one of two symmetrical differential representative vectors in the symmetrical relationship which can be selected as a differential representative vector for said next stage as it is, said generation and input means supplies said differential representative vector as it is as a differential representative vector of the next stage to said inner product computing means; and

in the case where said read differential representative vector is the other one of said two symmetrical differential representative vectors, said generation and input means multi plies said read differential representative vector by a negative number, and then supplies the result of multiplication as a differential representative vector into said inner product computing means in the next stage.

10. An image signal analyzing system of claim 1 wherein when inner products between mean separated block and said n stages of a differential representative vector set are computed by said inner product computing means, representative vectors constituting said differential representative vector set are defined to be vectors of a shape having a predetermined stepwise gradient, and values of elements of a required representative vector are approximated by a power of 2.

11. An image signal analyzing system of claim 2 wherein when inner products between mean separated block and said n stages of a differential representative vector set are computed by said inner product computing means, representative vectors constituting said differential representative vector set are defined to be vectors of a shape having a predetermined stepwise gradient, and values of elements of a required representative vector are approximated by a power of 2.