# US 2022/0044377 A1 · Method for reconstructing an image

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0044377 A1
- Document ID: US20220044377A1
- Status: Published application
- Kind: A1
- Application number: 17/125,301
- Filed: 2020-12-17
- Published: 2022-02-10
- Earliest claimed: 2020-08-04
- Assignees: V5 Technologies Co., Ltd.
- Applicants: V5 Technologies Co., Ltd.
- Inventors: Sheng-Chih HSU; Chien-Ting CHEN
- Web page: https://patentgrep.com/patent/US20220044377A1

## CPC

- G06T 7/001: {using an image reference approach}
- G06T 7/13: Edge detection
- G01N 21/9503: {Wafer edge inspection}
- G06T 3/10: Selection of transformation methods according to the characteristics of the input images
- G06T 7/149: involving deformable models, e.g. active contour models
- G06T 7/50: Depth or shape recovery
- G06T 2207/30148: Semiconductor; IC; Wafer
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)

## Abstract

A method for constructing an image includes: defining a foreground area associated with an object in an original image; identifying a plurality of contour points defining a contour of the object; for each of the contour points, obtaining a reference contour point set that includes at least one reference contour point on each of two sides of the contour point; obtaining a plurality of characteristic lines, each associated with the reference contour point set and defined by an end point obtained from the contour points; and aligning the end points on one side to form a straight edge and making the characteristic lines adjoin each other side by side, so as to construct a reconstructed image.

## Claims

### Claim 1

A method for constructing an image, the method to be implemented using a processor of an electronic device and comprising steps of:
a) obtaining an original image;
b) defining a foreground area that is associated with an object in the original image;
c) identifying a plurality of contour points that define a contour of the object;
d) for each of the plurality of contour points, obtaining a reference contour point set that includes, on each of two sides of the contour points, at least one reference contour point which is another one of the plurality of contour points;
e) obtaining a plurality of characteristic lines, each of the characteristic lines being defined by an end point obtained from the contour points, is associated with the corresponding reference contour point set, and has a pre-defined length and a predetermined width;
f) obtaining, for each of the characteristic lines, a plurality of pixel value sets that correspond respectively with the a plurality of pixels on the original image that constitute the characteristic line; and
g) rearranging the characteristic lines by aligning the end points on one side to form a straight edge and making the characteristic lines adjoin each other side by side, so as to construct a reconstructed image.

### Claim 2 (depends on 1)

The method of claim 1, wherein step e) includes, for each one of the plurality of contour points, sub-steps of:
e-1) for each of the reference contour points included in the reference contour point set that corresponds to said one of the plurality of contour point, obtaining a connecting line defined by the contour point and the reference contour point, and obtaining a reference length related to said one of the plurality of contour points and the reference contour point;
e-2) for each of the connecting lines obtained in sub-step e-1), obtaining a normal vector;
e-3) for each of the normal vectors obtained in sub-step e-2), calculating a reference angle between the normal vector and a predetermined line;
e-4) for each of the reference contour points, assigning a calculating weight for the reference contour point based on the corresponding reference length that is related to the reference contour point and said one of the plurality of contour points;
e-5) calculating a target angle based on the reference contour points included in the reference contour point set and the calculating weights assigned to the reference contour points; and
e-6) defining the characteristic line to pass through said one of the plurality of contour points, wherein an angle defined by the characteristic line and the predetermined line equals the target angle.

### Claim 3 (depends on 2)

The method of claim 2, wherein:
step d) includes, for each one of the plurality of contour points, obtaining a reference contour point set that includes, on each of the two sides of said one of the plurality of contour points, one reference contour point;
sub-step e-4) includes assigning calculating weights for the reference contour points using the following equations
W
S
=
d
EC
_
d
SC
_
+
d
EC
_
,
W
E
=
d
SC
_
d
SC
_
+
d
EC
_
,
where WS represents the calculating weight for one of the reference contour points and WE represents the calculating weight for the other one of the reference contour points, and dSC represents the reference length of said one of the reference contour points and dEC represents the reference length of said the other one of the reference contour points; and
sub-step e-5) includes calculating a target angle using the following equation
θC=WSθS+WEθE,
where θC represents the target angle, es represents the reference angle associated with said one of the reference contour points, and θE represents the reference angle associated with said the other one of the reference contour points.

### Claim 4 (depends on 3)

The method of claim 3, wherein:
step d) includes, for each one of the plurality of contour points, obtaining, on each of the two sides of said one of the plurality of contour points, one reference contour point that is immediately adjacent to said one of the plurality of contour points;
sub-step e-1) includes calculating the reference lengths related to said one of the plurality of contour points and the reference contour points using the following equations:
dSC=max(|xC−xS|,|yC−yS|),
dEC=max(|xC−xE|,|yC−yE|)
where (xC,yC) represents a coordinate set of said one of the plurality of contour points in a Cartesian coordinate system, (xS,yS) represents a coordinate set of one of the reference contour points included in the reference contour point set in the Cartesian coordinate system, and (xE,yE) represents a coordinate set of the other one of the reference contour points included in the reference contour point set in the Cartesian coordinate system.

### Claim 5 (depends on 2)

The method of claim 2, wherein:
step d) includes, for each one of the plurality of contour points, obtaining a reference contour point set that includes a number NL of reference contour points on one of the two sides of said one of the plurality of the contour points, and a number NR of reference contour points on the other one of the two sides of said one of the plurality of the contour points;
sub-step e-5) includes calculating a target angle using the following equations:
θC=stWL1θL1+stWL2θL2+ . . . +stWLNL-1θLNL-1+stWLNLθLNL+stWR1θR1+stWR2θR2+ . . . +stWRNR-1θRNR-1+stWRNRθRNR
where θC represents the target angle, stWLn represents the calculating weight for an nth one of the reference contour points on one of the two sides, stWRm represents the calculating weight for an mth one of the reference contour points on the other one of the two sides, θLn represents the reference angle associated with an nth one of the reference contour points on said one of the two sides, and θRm represents the reference angle associated with an mth one of the reference contour points on said the other one of the two sides, where n=1, 2, . . . NL−1, and NL, and m=1, 2, . . . NR−1, and NR.

### Claim 6 (depends on 5)

The method of claim 5, wherein the number NL is equal to the number NR, and is an even number greater than 2.

### Claim 7 (depends on 1)

The method of claim 1, wherein in step e), each of the characteristic lines has the corresponding one of the contour points serving as the end point, and has a length that is not less than a greatest distances among distances each between a centroid of the object and one of the contour points.

### Claim 8 (depends on 7)

The method of claim 7, wherein in step e), the characteristic lines have a uniform length.

### Claim 9 (depends on 1)

The method of claim 1, further comprising, between steps c) and d):
performing a curve fitting operation to construct a fitted curve using a centroid of the object, the contour points and a fitting function, the fitted curve being composed of a plurality of curve points; and
performing an expanding operation on the fitted curve to obtain an expanded curve that is composed of a plurality of expanded curve points, wherein each of the expanded curve points is a point radially spaced apart from a corresponding one of the curve points in a direction away from the centroid of the object by a predetermined expanding distance, and each of the contour points is encompassed within the expanded curve;
wherein for each of the characteristic lines, the end point is one of the expanded curve points and the characteristic line passes through a corresponding one of the curve points.

### Claim 10 (depends on 9)

The method of claim 9, the object being a semiconductor wafer and having an approximately elliptical shape, wherein the fitting function is an ellipse function expressing a standard ellipse, and the fitted curve is elliptical.

### Claim 11 (depends on 1)

The method of claim 1, wherein step c) includes performing binarization on the original image.
