Image processor and method thereof

• US 7,283,679 B2
• Filed: 06/11/2003
• Issued: 10/16/2007
• Est. Priority Date: 06/12/2002
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Original Assignee: Kabushiki Kaisha Toshiba (Toshiba Corporation)
• Inventors: Okada, Ryuzo, Miyazaki, Hiroshi
• Primary Examiner(s): Poon; King Y.
• Assistant Examiner(s): Vo; Quang N.
• Application Number:
  US10/458,265
• Publication Number:
  US 20040032990A1
• Time in Patent Office: 1,588 Days
• Field of Search: 358/3.26, 382/288, 382/260, 382/274, 382/299, 382/300, 348/222
• US Class Current: 382/260
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/94 based on local image properties, e.g. for local contrast enhancement

Abstract

An image-processor for shooting images at a high frame rate in time series maintaining a high sensitivity and a high speed is provided. The image-processor comprises a high-speed image-obtaining unit 1 for arranging the imaging elements in a one-dimensional or two-dimensional manner to obtain images at a high frame rate, a pixel selection unit 2 for selecting pixels constituting the image maintaining a predetermined sampling interval, and a peripheral pixel value addition unit 3 for a adding up pixel values of pixels selected by the pixel selection unit 2 and pixel values of peripheral pixeling. It comprises further a pixel value adjusting unit 4 for adjusting the pixel values of the sample pixels calculated by the peripheral pixel value addition unit 3 so as to be suited as pixel values of the output image, and an image output unit 5 for producing pixel values of sample pixels calculated by the pixel value adjusting unit 4 as an image of one frame.

Claims

1. An image processor which processes images shot in time series to improve sensitivity and to reduce noise, the image processor comprising:

pixel selection means for selecting a plurality of sample pixels at a predetermined sampling interval from pixels of the images shot in time series;

pixel combining means for calculating a weighted sum of peripheral pixel values with respect to each sample pixel by using weights compliant with a Gaussian function, and for computing an enhanced pixel value of each sample pixel based on the calculated weighted sum so as to improve the sensitivity;

pixel value adjusting means for adjusting the enhanced pixel values of the sample pixels to a predetermined range of pixel values; and

image output means for producing an output image including the adjusted pixel values of the sample pixels;

wherein the pixel value adjusting means adjusts the enhanced pixel values of the sample pixels to the predetermined range either by executing a first process of multiplying the enhanced pixel values of the sample pixels by a predetermined constant, or by executing a second process of replacing the pixel values by a maximum value when the pixel values exceed the maximum value, or by executing both the first and the second process; and

wherein the sampling interval of the sample pixels is narrower than a range of the peripheral pixels.

2. An image processor according to claim 1, wherein

the images of time series may be high-speed images shot at 100 frames/second to 2000 frames/second.

3. An image processor according to claim 1, wherein

the images of time series may be images of time series around a moving body which is shot by an image obtaining means mounted on the moving body, and the images output from the image output means may be sent to a control unit in the moving body.

4. A method which processes a time series of images to improve sensitivity and to reduce noise, the method comprising the steps of:

selecting a plurality of sample pixels at a predetermined sampling interval from pixels of the images in time series;

calculating a weighted sum of peripheral pixel values with respect to each sample pixel by using weights compliant with a Gaussian function, and computing an enhanced pixel value of each sample pixel based on the calculated weighted sum so as to improve the sensitivity;

adjusting the enhanced pixel values of the sample pixels to a predetermined range of pixel values; and

producing an output image including the adjusted pixel values of the sample pixels;

wherein the enhanced pixel values of the sample pixels are adjusted to the predetermined range either by executing a first process of multiplying the enhanced pixel values of the sample pixels by a predetermined constant, or by executing a second process of replacing the pixel values by a maximum value when the pixel values exceed the maximum value, or by executing both of the first and the second processes; and

wherein the sampling interval of the sample pixels is narrower than a range of the peripheral pixels.

5. A computer readable medium including a program for causing a computer to perform a method for processing a time series of images, the method comprising:

selecting a plurality of sample pixels at a predetermined sampling interval from pixels of the images in the time series;

calculating a weighted sum of peripheral pixel values with respect to each sample pixel by using weights compliant with a Gaussian function, and computing an enhanced pixel value of each sample pixel based on the calculated weighted sum so as to improve the sensitivity;

adjusting the enhanced pixel values of the sample pixels to a predetermined range of pixel values; and

producing an output image including the adjusted pixel values of the sample pixels;

wherein the enhanced pixel values of the sample pixels are adjusted to the predetermined range either by executing a first process of multiplying the enhanced pixel values of the sample pixels by a predetermined constant, or by executing a second process of replacing the pixel values by a maximum value when the pixel values exceed the maximum value, or by executing both of the first and the second processes; and

wherein the sampling interval of the sample pixels is narrower than a range of the peripheral pixels.

6. An image processor, comprising:

an image obtaining device to obtain images;

a pixel selection unit configured to select a plurality of sample pixels at a predetermined sampling interval from the pixels of the images;

a calculating unit configured to calculate a weighted sum of peripheral pixels with the sample pixels as centers by using a weight compliant with a Gaussian function, and configured to compute an enhanced pixel value of each sample pixel based on the calculated weighted sum so as to improve sensitivity;

an adjusting unit configured to adjust the enhanced pixel values of the sample pixels to a predetermined range of pixel values; and

an image output device to output an output image having the adjusted pixel values of the sample pixels;

wherein the adjusting unit adjusts the enhanced pixel values of the sample pixels to the predetermined range either by executing a first process of multiplying the enhanced pixel values of the sample pixels by a predetermined constant, or by executing a second process of replacing the pixel values by a maximum value when the pixel values exceed the maximum value, or by executing both of the first and the second processes; and

wherein the sampling interval of the sample pixels is narrower than a range of the peripheral pixels.

7. An image processor according to claim 6, wherein

the images of time series are high-speed images shot at 100 frames/second to 2000 frames/second.

8. An image processor according to claim 6, wherein

the images of time series may be images of time series around a moving body which is shot by an image obtaining means mounted on the moving body, and the images output from the image output means may be sent to a control unit in the moving body.