Image data processor

• US 4,254,400 A
• Filed: 11/28/1979
• Issued: 03/03/1981
• Est. Priority Date: 12/13/1978
• Status: Expired due to Term

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Motoike, Jun, Yoda, Haruo
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US06/098,074
• Time in Patent Office: 461 Days
• Field of Search: 340/146.3 MA, 340/146.3 AE, 340/146.3 Y, 340/146.3 S, 340/146.3 Q, 340/146.3 AH, 340/146.3 AC, 358/105, 358/106, 358/107
• US Class Current: 382/194
• CPC Class Codes: B66B 1/3476 Load weighing or car passenger counting devices B66B5/14 takes precedence
• CPC Class Codes: G06V 10/40 Extraction of image or video features
• CPC Class Codes: G08B 13/19602 Image analysis to detect motion of the intruder, e.g. by frame subtraction
• CPC Class Codes: G08B 13/19634 Electrical details of the system, e.g. component blocks for carrying out specific functions
• CPC Class Codes: G08B 13/19676 Temporary storage, e.g. cyclic memory, buffer storage on pre-alarm

Abstract

An image data processor for recognizing the degree of variation in the object in the image by dividing an image plane into a plurality of block regions and by sequentially comparing an input image and a background image for each of the divided block regions. The data to be used for comparing the input image and the background image are either the number of occurences of preset amplitude relationships of the signals which are produced by extracting the signals of plural pairs of picture elements in spacially preset relative position relationships from the output signals of an image pickup device while sequentially shifting the positions of the picture elements and by accomplishing the counting operations for each pair of the picture elements when the picture element extracted is included in a preset block region or the data which are produced by linearly transforming the data of the number of occurences.

Claims

1. An image data processor comprising:

first means (1 and

     2) for converting an image including an object to be recognized into spacially quantitized picture element signals;

second means (3) for extracting plural pairs of the picture element signals, which are in spacially preset relative position relationships, from the output of the first-named means while sequentially shifting the picture element positions to be extracted;

third means (4) connected with the second-named means for counting the number of occurences of preset amplitude relationships between the plural pairs of the input signals when the output signals from the second-named means indicate the picture elements which are included in a preset region of the image;

fourth means (7) for dividing a reference image into a plurality of regions to store in advance the feature data of each of the divided regions; and

fifth means (8) for sequencially comparing the counted data generated from the third-named means and those data of the fourth-named means, which correspond to said preset regions,

whereby the object in the image is recognized from the compared results of the fifth-named means.

3. An image data processor as set forth in claim 1 or 2, wherein the third-named means includes:

a plurality of comparators made receptive of a pair of the picture element signals from the second-named means for generating signals when a preset amplitude relationship resides among the input signals;

a plurality of counters provided for said comparators, respectively, for accomplishing their counting operations in accordance with the output signals of said comparators, respectively; and

means for presetting the operating periods of said counters, respectively.

2. An image data processor comprising:

first means (1 and

     2) for converting an image including an object to be recognized into spacially quantitized picture element signals;

second means (3) for extracting plural pairs of the picture element signals, which are in spacially preset relative position relationships, from the output of the first-named means while sequentially shifting the picture element positions to be extracted;

third means (4) connected with the second-named means for counting the number of occurences of preset amplitude relationships between the plural pairs of the input signals when the output signals from the second-named means indicate the picture elements which are included in a preset region of the image;

linear transformation means (6) for converting the relationships between the counted data generated from the third-named means and the paired picture elements corresponding thereto into one-dimentional data;

fourth means (7) for dividing a reference image into a plurality of regions to store in advance the feature data of each of the divided regions; and

fifth means (8) for sequentially comparing the data generated from said linear transformation means and those data of the fourth-named means, which correspond to said preset regions,

whereby the object in the image is recognized from the compared results of the fifth-named means.