Adaptive quantization compatible with the JPEG baseline sequential mode

• US 6,175,650 B1
• Filed: 01/26/1998
• Issued: 01/16/2001
• Est. Priority Date: 01/26/1998
• Status: Expired due to Term

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Sindhu, Pradeep S., Kletter, Doron, Frailong, Jean-Marc, Nafarieh, Asghar, Curry, Donald J.
• Primary Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US09/013,480
• Time in Patent Office: 1,086 Days
• Field of Search: 382/171, 382/164, 382/236, 382/239, 382/108, 382/172, 382/173, 382/251, 358/453, 358/430, 358/462, 375/240.03, 375/240.02
• US Class Current: 382/171
• CPC Class Codes: G06T 9/005 Statistical coding, e.g. Huffman, run length coding

Abstract

Pixel blocks of an input image are type classified based on an analysis of the secord differences between the values of neighboring pixels in the rows and columns of each pixel block. A histogram of these second order differences may be employed to refine this analysis

Claims

1. A method for type characterizing a pixel block representation of an image region, said pixel block including a plurality of pixel rows and a plurality of pixel columns;

said method comprising;

inputting image signal data representing values for pixels in the pixel rows and pixel columns of the pixel block; and

obtaining spatial domain texture values for the respective rows and columns of said pixel block based on thresholded second order differences between the values of neighboring pixels in said rows and columns.

2. The method of claim 1, further comprising:

calculating histogram values to determine a numerical distribution of the texture values obtained for the pixel rows and pixel columns; and

determining the image type based on the distribution of the texture values.

3. A method for type characterizing a pixel block representation of a region of an image, said pixel block having a plurality of rows and a plurality of columns, said method comprising:

inputting image signal data representing values for pixels in the rows and columns of the pixel block;

calculating quantized first order differences between the values for neighboring pixels in said rows and columns using quantization thresholds that are selected to reject noise in said pixel values;

obtaining texture values along each of the rows and columns of said block by a second order spatial domain analysis of said quantized first order differences between the values of said neighboring pixels;

calculating histogram values to determine a numerical distribution of the texture values obtained for the pixel rows and pixel columns; and

determining the image type based on the distribution of the texture values.

4. A segmenter comprising:

a pixel block input for receiving a pixel block representation of a region of an image;

an image texture module that determines an image type of said pixel block of an image by calculating spatial domain texture values based on a second order differential analysis of quantized first order differences between values for neighboring pixels in rows and columns of the pixel block;

a pixel block type output for outputting said determined image type.

5. The segmenter of claim 4, wherein the image texture module further calculates histogram values to determine a numerical distribution of the texture values obtained for the pixel rows and pixel columns, and determines the image type based on the distribution of the texture values.

6. A segmenter comprising:

a pixel block input for receiving a pixel block representation of a region of an image;

an image texture module that determines an image type of said pixel block of an image by calculating spatial domain texture values along a plurality of pixel rows and pixel columns of the pixel block based on thresholded second order differences between the values of neighboring pixels in said rows and columns, calculates histogram values to determine a numerical distribution of the texture values obtained along the pixel rows and pixel columns, and determines the image type based on the distribution of the texture values;

a pixel block type output for outputting said determined image type.