Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

IMAGE ENHANCEMENT FILTER

IP.com Disclosure Number: IPCOM000024315D
Original Publication Date: 1980-Apr-30
Included in the Prior Art Database: 2004-Apr-02
Document File: 2 page(s) / 86K

Publishing Venue

Xerox Disclosure Journal

Abstract

This disclosure is to report an image enhancement filter which is a linear combination of an unsharp masking filter and a maximum entropy filter (Figure 1). According to the arrangement shown in Figure 1, the enhanced image may be expressed as:

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 83% of the total text.

Page 1 of 2

IMAGE ENHANCEMENT FILTER
235/156 C1. U.S. Proposed H. S. Hou Classification

Int. C1. G06f 7/38

MAXIMUM ENTROPHY FILTER

fi

$N_SH_ARP MA_SKyG- I

IT I

HOLD CKT. L---J

ANALOG INPUT

ALL OPAMPS ARE No. SN 52733

me. 2

Volume 5 Number 2 March/April 1980 173

[This page contains 1 picture or other non-text object]

Page 2 of 2

IMAGE ENHANCEMENT FILTER (Contld)

This disclosure is to report an image enhancement filter which is a linear combination of an unsharp masking filter and a maximum entropy filter (Figure 1). According to the arrangement shown in Figure 1, the enhanced image may be expressed as:

A N

gi = kfi + (1-k)fi

where k is a gain factor ranged from 0.5 to 0.6 and the f's and the g with subscript i represent the filtered and enhanced image samples at position i.

The unsharp masking operaticn is to crispen the edges of an image. The pixel value of the edge enhanced image, fi is related to the original pixel value, fi by

- 0.0095 (fi+4 + fi-4).

The maximum entropy filter which is a kind of histogram equalization procedure is to enhance the contrast of an image. It was derived from the consideration of not only resulting maximum entropy but also retaining the local greyness and the relative pixel position. The filtered image from the maximum entropy filter may be expressed as:

rv 2 2
f. = MiSi /Mi + Si)

1

where Si is the sum of the local image pixels, i.e.,

si = fi-4 + fi-3 + fi,2 + fi-l + fi + + fi+2 + fi+3 + fi+4,

and Mi is a low-passed image of pixels fi,4 to f i+...