Arrangement and method for processing image data

• US 7,327,856 B2
• Filed: 05/25/2001
• Issued: 02/05/2008
• Est. Priority Date: 05/25/2001
• Status: Expired due to Fees

• Current Assignee: Continental Automotive GmbH (Continental AG)
• Original Assignee: Siemens AG
• Inventors: Wagner, Ulrich, Rosl, Reinhard, Spalke, Stefan, Belau, Horst
• Primary Examiner(s): Ahmed; Samir
• Assistant Examiner(s): Tabatabai; Abolfazl
• Application Number:
  US10/478,510
• Publication Number:
  US 20040234100A1
• Time in Patent Office: 2,447 Days
• Field of Search: 382/100, 382/104, 382/106, 280/732, 280/735, 340/436, 340/438, 340/901, 340/903, 701/45, 701/301
• US Class Current: 382/106
• CPC Class Codes: G06T 5/80 Geometric correction

Abstract

A source image (S1) distorted by a camera lens can be transformed into a rectified target image (T), by means of a tabular imaging specification. The above occurs during read-out from the image sensor and in real-time. Each source pixel in the source image is assigned none, one or several target pixels in the target image (T1). A first controller (C1) controls the image sensors (B1, B2) accurately with tine and the image equalization and image correlation. A second controller (C2) controls the first controller (C1) and works in a manner temporally decoupled from the above.

Claims

1. An arrangement for processing image data in imaging systems for occupant protection systems, comprising:

a first control unit which, in real time, reads out from at least one image sensor, grayscale values of the image data as source image pixels;

an optical system which distorts the image data, where the first control unit corrects the image data distorted by the optical system using a tabular mapping rule so that a target image is produced and provides range information of the target image; and

a second control unit for controlling the first control unit and taking over the target image data and the range information of the target image processed by the first control unit, the second control unit performing seat occupancy classification, wherein

by mapping the source image into the target image according to the mapping rule, the target image having fewer pixels than the source image can be obtained.

2. The arrangement according to claim 1, wherein

the first control unit contains a rectification table stored in a first memory and which can be used to execute the mapping rule for correcting the image supplied by the image sensor.

3. The arrangement according to claim 2, wherein

to initialize the rectification table, image data of a reference image captured by one or more image sensors is stored in the first memory.

4. The arrangement according to claim 1, wherein

the first control unit is implemented as an ASIC or FPGA,

the second control unit is implemented as a microcontroller or microprocessor, and

the first and the second control units are disposed separately or one of in an IC package andon a chip.

5. The arrangement according to claim 1, wherein

two image sensors are present, the first control unit correlating the corrected image data in each case to produce a range image.

6. The arrangement according to claim 1, wherein

at least one image sensor is linked via the first control unit to the second control unit.

7. A method for processing image data in imaging systems for occupant protection systems, comprising:

reading the image data available in an image sensor of a source image distorted by an optical system from the image sensor by a first control unit in which the image data is corrected in real time using a tabular mapping rule so that a target image is produced and is provided with range information, whereby a seat occupancy classification is carried out using in a second control unit which controls the first control unit, the corrected target image data and range information taken from the first control unit is provided with range information and classified,

wherein

the target image has fewer pixels than the source image.

8. The method according to claim 7, wherein

a pixel under a predefined pixel address contains a grayscale value.

9. The method according to claim 7, wherein

the target image or the image data of the target image is buffered in a memory of the first control unit.

10. The method according to claim 7, wherein

if a first and a second image sensor are present, the two corrected target images produced are correlated row-wise with one another to determine range information,

the resulting image pixels or image data of the image pixels provided with disposed in at least one image row, are buffered in a memory of the first control unit, and

at the request of a second control unit, the buffered image pixels or image data of the image pixels are transferred to the second control unit for further processing, vehicle seat occupancy classification being performed in the second control unit.

11. The method according to claim 10, wherein

the result of the vehicle seat occupancy classification is forwarded to a central occupant protection control unit.

12. The method according to claim 10, wherein

a predefined reference image is captured by at least one image sensor to initialize or create a rectification table,

the captured reference source image is read out of the image sensor and stored in a first memory,

the rectification table is created from the stored reference source image, and

the created rectification table is stored in the first memory, whereby

the reference source image stored in the first memory is at least partially overwritten by the rectification table.

13. The method according to claim 12, wherein

the reference source image has a predefined pattern.

14. The method according to claim 12, wherein

the reference source image is captured by the two image sensors and used as reference for the correlation.

15. The method according to claim 12, wherein

a rectification table is provided for each image sensor or one rectification table is available for all the image sensors.

16. The method according to claim 7, wherein

a system clock of the second control unit is essentially independent of a system clock of the first control unit.

17. The method according to claim 7, wherein

image processing in the first control unit is executed in parallel with classification in the second control unit.

18. The method according to claim 7, wherein

the first and the second control unit operate largely independently of one another.