# US 2017/0255807 A1 · Method for improving images applicable to fingerprint images

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

- Number: US 2017/0255807 A1
- Document ID: US20170255807A1
- Status: Published application
- Kind: A1
- Application number: 15/451,246
- Filed: 2017-03-06
- Published: 2017-09-07
- Earliest claimed: 2016-03-04
- Assignees: Safran Identity & Security
- Applicants: Safran Identity & Security
- Inventors: Laurent KAZDAGHLI; Cedric THUILLIER; Lauriane COUTURIER
- Web page: https://patentgrep.com/patent/US20170255807A1

## CPC

- G06T 5/90: Dynamic range modification of images or parts thereof
- G06T 7/11: Region-based segmentation
- G06T 7/155: involving morphological operators
- G06V 10/431: {Frequency domain transformation; Autocorrelation}
- G06V 40/1306: {non-optical, e.g. ultrasonic or capacitive sensing}
- G06V 40/1365: {Matching; Classification}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

The invention relates to a method for processing an image comprising a set of pixels, each pixel being associated with a grey level, the method comprising a step of segmentation of the image to generate a modified image only containing regions of the image having an alternation of bright zones and dark zones at a frequency greater than a minimum frequency, said step of segmentation comprising: the allocation, to each pixel of the image, of a frequency response level, corresponding to a frequency of alternations of bright zones and dark zones in the vicinity of the pixel, the definition of regions of the image by grouping together neighbouring pixels of same frequency response level, the determination of a threshold frequency response level, and the generation of an image only comprising regions of which the pixels have a frequency response level greater than or equal to the threshold frequency response level.

## Claims

### Claim 1

Method for processing a finger image acquired by a fingerprint sensor of direct view type and comprising at least one fingerprint in order to implement an authentication or identification of an individual by comparison of fingerprints, said image comprising a set of pixels, each pixel being associated with a grey level, the method comprising a step of segmentation of the image to generate a modified image only containing regions of the image having an alternation of bright zones and dark zones at a frequency greater than a minimum frequency, said step of segmentation comprising:
the allocation, to each pixel of the image, of a frequency response level, corresponding to a frequency of alternations of bright zones and dark zones in the vicinity of the pixel, by evaluating the variabilities of the grey levels of the pixels contained in a window of determined size positioned around each pixel,
the definition of regions of the image by grouping together neighbouring pixels of same frequency response level previously allocated to each pixel,
the determination of a threshold frequency response level,
selection, from the regions of the image defined previously by grouping together neighbouring pixels of same frequency response level, of regions of which the pixels have a frequency response level greater than or equal to the threshold frequency response level, and
the generation of a modified image only comprising, among the regions of the image defined previously by grouping together neighbouring pixels of same frequency response level, regions of which the pixels have a frequency response level greater than or equal to the threshold frequency response level,
a step of authentication or identification of an individual by comparison of fingerprints on the basis of the modified image.

### Claim 2 (depends on 1)

Method according to claim 1, in which the allocation, to a pixel, of a frequency response level by evaluating the variabilities of the grey levels of the pixels contained in a window of determined size positioned around each pixel comprises the determination of a gradient of grey levels, the frequency response level allocated to said pixel being based on said gradient of grey levels.

### Claim 3 (depends on 1)

Method according to claim 1, in which the regions of the image defined by grouping together neighbouring pixels of same frequency response level are structured, to determine the threshold frequency response level, into a topological tree or a connected component tree, defining a parent relationship between said regions of the image.

### Claim 4 (depends on 1)

Method according to claim 1, in which the regions of the image are structured according to a parent-daughter relationship defined as follows, for each region i of the image:
initially all the neighbouring regions of the region i on the image are considered as potential parents,
if the number of potential parents is equal to 1, the neighbour j is assigned as parent to the region i, and the region i is removed from the list of potential parents of the region j,
if not, each parent region of the region i is selected from the list of potential parents having a frequency response level lower than that of the region i, as that having the level the closest to that of the region i, and the region i is removed from the list of potential parents of its parent region, and
if all the potential parents of the region i have a frequency response level greater than that of the region i, each parent region of the region i is determined as that having the frequency response level the closest to that of the region i and the region i is removed from the list of potential parents of its parent region.

### Claim 5 (depends on 3)

Method according to claim 3, in which the determination of the threshold frequency response level comprises the implementation of the following steps:
for a set frequency response level N, definition of a set of macro-regions designated level N, such that each macro-region of level N comprises a parent region of pixels of frequency response level lower than or equal to N and the set of daughter regions of this region,
for each value of frequency response level from an initial value of frequency response level in the image, calculation of the relative variation in surface of the macro-regions of a level Nn compared to the macro-regions of the preceding level Nn-1, and
the minimum frequency response level is determined as the level Ni for which the relative variation in surface of the macro-regions of level Ni compared to the preceding level Ni-1 is minimal.

### Claim 6 (depends on 1)

Method according to claim 1, in which the step of segmentation comprises, before the definition of the regions of the image by grouping together neighbouring pixels of same frequency response level, the implementation of the steps of:
morphological erosion then dilation of the image as a function of the values of grey level of the pixels,
generation of a differential image by subtraction, from the initial image, of the image having undergone morphological erosion and dilation,
application, to the differential image, of a median filter on the value of the frequency response level,
and in which the definition of the regions of the image by grouping together neighbouring pixels of same frequency response level is implemented on the image resulting from the application of the median filter on the differential image.

### Claim 7 (depends on 1)

Method according to claim 1, further comprising, after the step of segmentation, a step of enhancement of the grey levels associated with the pixels of the modified image, the degree of enhancement of the pixels of a region being a function of the frequency response level of the pixels of the region.

### Claim 8 (depends on 7)

Method according to claim 7, in which the higher the frequency response level of the pixels of a region of the modified image, the higher the degree of enhancement of grey levels associated with the pixels of the region.

### Claim 9 (depends on 1)

Computer programme product, comprising a computer readable support with a memory storing code instructions for the implementation of the method according to claim 1, when instructions are executed by a computer.

### Claim 10 (depends on 1)

Image processing system, comprising a processing unit of the processing means suited to implementing the method according to claim 1.

### Claim 11 (depends on 10)

Image processing system according to claim 10, further comprising an image acquisition means, said image acquisition means being a fingerprint sensor of thin film transistor type.
