SYSTEM AND METHOD FOR FILTERING IMAGE NOISE

• US 20070003158A1
• Filed: 05/29/2006
• Published: 01/04/2007
• Est. Priority Date: 07/01/2005
• Status: Active Grant

• Current Assignee: Hong Fu Jin Precision Industry (Shenzhen) Co., Ltd. (Hon Hai Precision Industry Co., Ltd.), Hon Hai Precision Industry Co., Ltd.
• Original Assignee: Hon Hai Precision Industry Co., Ltd.
• Inventors: CHANG, CHIH-KUANG, Jiang, Li, Sun, Xiao-Chao
• Granted Patent: US 7,697,780 B2
• Time in Patent Office: Days
• US Class Current: 382/275
• CPC Class Codes: G06T 2207/20192 Edge enhancement; Edge preservation
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening

Abstract

A method for filtering image noise includes the steps of: acquiring an image; converting gray-scale values of the image and sharpening the image; defining axes perpendicular to image borders and selecting intersecting points between the axes and the image border; creating an object points set and inputting all the intersecting points into the object points set; fitting a geometric character according to the object points set; defining a tolerance threshold value; selecting an intersecting point from the object points set; calculating a distance from the selected intersecting point to the geometric character; determining whether the distance is larger than the defined tolerance threshold value; marking the selected intersecting point as a noise if the distance is larger than the defined tolerance threshold value; and deleting the selected intersecting point from the object points set. A related method is also disclosed.

Claims

1. A system for filtering image noise, the system comprising:

an image acquiring module for acquiring an image;

a gray-scale converter for converting gray-scale values of the image, and for sharpening the image;

an intersecting point selecting module for defining axes perpendicular to image borders, for determining gray-scale values of axis points, for selecting intersecting points between the image borders and the axes, for creating an object points set, and for inputting the intersecting points into the object points set; and

a noise filtering module for fitting a geometric character according to the object points set and defining a tolerance threshold value, for calculating a distance from each intersecting point in the object points set to the geometric character, for determining whether the distance is larger than the defined tolerance threshold value, and for marking the intersecting point as a noise and deleting the intersecting point if the distance is larger than the defined tolerance threshold value.

2. The system as claimed in claim 1, wherein the system further comprises a storing module for storing different kinds of data.

3. The system as claimed in claim 1, wherein the gray-scale converter is further used for counting a frequency of each gray-scale value.

4. The system as claimed in claim 1, wherein the gray-scale converter is further used for calculating a distribution range of the gray-scale values and defining a changing range.

5. A method for filtering image noise, the method comprising the steps of:

acquiring an image;

converting gray-scale values of the image and sharpening the image;

defining axes perpendicular to image borders and selecting intersecting points between the axes and the image border;

creating an object points set and inputting all the intersecting points into the object points set;

fitting a geometric character according to the object points set;

defining a tolerance threshold value;

selecting an intersecting point from the object points set;

calculating a distance from the selected intersecting point to the geometric character;

determining whether the distance is larger than the defined tolerance threshold value;

marking the selected intersecting point as a noise if the distance is larger than the defined tolerance threshold value; and

deleting the selected intersecting point from the object points set.

6. The method according to claim 5, wherein the geometric character is a line, and the tolerance threshold value is a tenth of the length of the line.

7. The method according to claim 5, wherein the geometric character is one of a circle or an arc, the tolerance threshold value is a tenth of the diameter of the one of the circle or the arc.

8. The method according to claim 5, further comprising the steps of:

determining whether all the intersecting points in the object points set have been selected;

creating a new object points set if all the intersecting points in the object points set have been selected;

fitting a new geometric character according to the new object points set; and

outputting and storing the new geometric character.

9. The method according to claim 8, wherein the step of determining whether all intersecting points in the object points set have been selected comprises the step of returning to the selecting step for selecting another intersecting point in the object points set if there are intersecting points have not been selected.

10. The method according to claim 8, wherein the step of determining whether the distance is larger than the defined tolerance threshold value comprises the step of executing the step of determining whether all intersecting points in the object points set have been selected if the distance is smaller than the defined tolerance threshold value.

11. The method according to claim 5, wherein the converting step comprises the steps of:

calculating a size of the image by reading the length and the width of the image;

counting a frequency of each gray-scale value of the image;

calculating a distribution range of the gray-scale values of the image;

defining a changing range of the gray-scale values;

converting the gray-scale values of the image according to a particular linear changing formula;

counting a new frequency of each converted gray-scale value;

determining whether the distribution of the converted gray-scale values is regular according to the changing range; and

sharpening the image if the distribution of the converted gray-scale values is regular.

12. The method according to claim 11, wherein the determining step comprises the step of returning to the step of defining a changing range of the gray-scale values, if the distribution of the converted gray-scale values is irregular.

13. The method according to claim 5, wherein the steps of defining axes perpendicular to image borders and selecting intersecting points between the axes and the image border comprises the steps of:

defining an axis perpendicular to image borders;

determining coordinates of each axis point;

determining a gray-scale value of each axis point according to an average value filtering method;

selecting an intersecting point of the image according to a comparing method;

determining whether the axis is the last one that can be defined to be perpendicular to the image borders; and

storing the object points set if the axis is the last one.

14. The method according to claim 13, further comprising the step of:

returning to the step of defining an axis perpendicular to image borders, if the axis is not the last one.

15. The method according to claim 13, wherein the average value filtering method comprises the steps of:

selecting an axis point;

determining the coordinates and the gray-scale value of the axis point;

selecting a plurality of nearby points that around the axis point;

determining the corresponding gray-scale values of the plurality of nearby points;

calculating an average value of the gray-scale values of the plurality of nearby points;

determining absolute differences between the average value and each gray-scale value of the plurality of nearby points respectively;

defining a threshold value, the threshold value being not less than 0;

comparing the absolute differences and the threshold value respectively; and

using the average value to replace corresponding gray-scale value if the threshold value is larger than the absolute difference, or keeping current gray-scale value if the threshold value is smaller than the absolute difference.

16. The method according to claim 15, wherein the plurality of nearby points are arranged in a symmetrical area.

17. The method according to claim 15, wherein the definition of the threshold value is determined according to one-tenth of a difference between the distribution ranges.

18. The method according to claim 13, wherein the comparing method comprises the steps of:

comparing gray-scale values of every two continuous axis points;

acquiring compared results;

selecting a largest compared result;

determining the two continuous axis points correspond to the largest compared result; and

selecting the latter axis point that has the largest gray-scale value compared result as the intersecting point from the two continuous axis points.

19. A system for filtering image noise, the system comprising:

an image acquiring module for acquiring an image;

a gray-scale converter for converting gray-scale values of the image;

an intersecting point selecting module for defining axes perpendicular to image borders, for selecting intersecting points between the image borders and the axes, for creating an object points set, and for inputting the intersecting points into the object points set; and

a noise filtering module for fitting a geometric character according to the object points set.