# US 2026/0203968 A1 · Palm print image generation method

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

- Number: US 2026/0203968 A1
- Document ID: US20260203968A1
- Status: Published application
- Kind: A1
- Application number: 19/559,274
- Filed: 2026-03-06
- Published: 2026-07-16
- Earliest claimed: 2023-11-17
- Assignees: Tencent Technology (Shenzhen) Company Limited
- Applicants: Tencent Technology (Shenzhen) Company Limited
- Inventors: Lei SHEN; Jianlong JIN; Ruixin ZHANG; Jingyun ZHANG; Shouhong DING
- Web page: https://patentgrep.com/patent/US20260203968A1

## CPC

- G06N 3/045: Combinations of networks
- G06T 11/10: Texturing; Colouring; Generation of textures or colours (retouching, inpainting or scratch removal G06T5/77)
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06T 5/73: Deblurring; Sharpening
- G06T 11/23: using straight lines or curves
- G06V 10/761: {Proximity, similarity or dissimilarity measures}
- G06V 10/772: Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
- G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
- G06V 10/82: using neural networks
- G06V 40/1359: {Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop}
- G06T 2207/20081: Training; Learning
- G06T 2207/20192: Edge enhancement; Edge preservation
- G06T 2207/30196: Human being; Person
- G06V 10/54: relating to texture
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

In a palm print image generation method, a plurality of control points is determined based on a palm print curve template, and a set of Bezier curves is generated based on the plurality of control points. In the method, feature extraction is performed on the set of Bezier curves, to obtain a feature map of the set of Bezier curves. In the method, a palm crease energy image that includes a set of palm creases is generated based on the feature map, the set of palm creases having a line distribution corresponding to a line distribution of the set of Bezier curves and having line stroke properties corresponding to line orientation energy of each pixel on the set of palm creases. In the method, a simulated palm print image having texture information is generated based on the palm crease energy image.

## Claims

### Claim 1

A palm print image generation method, comprising:
determining a plurality of control points based on a palm print curve template;
generating a set of Bezier curves based on the plurality of control points;
performing, by processing circuitry, feature extraction on the set of Bezier curves, to obtain a feature map of the set of Bezier curves;
generating a palm crease energy image that includes a set of palm creases based on the feature map, the set of palm creases having a line distribution corresponding to a line distribution of the set of Bezier curves and having line stroke properties corresponding to line orientation energy of each pixel on the set of palm creases; and
generating, by the processing circuitry, a simulated palm print image having texture information based on the palm crease energy image.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the generating the palm crease energy image comprises:
determining a line energy feature of each pixel in each per-channel feature map of the feature map that is a multi-channel feature map;
obtaining an enhanced multi-channel feature map based on enhancing the line energy feature of each pixel in each per-channel feature map of the feature map; and
generating the palm crease energy image based on the enhanced multi-channel feature map.

### Claim 3 (depends on 2)

The method according to claim 2, wherein the determining the line energy feature of each pixel comprises, for each per-channel feature map:
extracting line orientation energy components of each pixel from the per-channel feature map by using a linear convolution layer, the linear convolution layer including a respective Gaussian modified finite Radon transform core in each of a plurality of pre-determined directions, and the line orientation energy components corresponding to line orientation energy of the pixel in the plurality of pre-determined directions; and
obtaining the line energy feature of each pixel based on a maximum line orientation energy component of the line orientation energy components and one of the plurality of pre-determined directions corresponding to the maximum line orientation energy component.

### Claim 4 (depends on 1)

The method according to claim 1, wherein the generating the simulated palm print image comprises:
generating the simulated palm print image based on the palm crease energy image and one or more control vectors configured to increase diversity among generated images.

### Claim 5 (depends on 1)

The method according to claim 1, wherein the set of palm creases included in the palm crease energy image indicates an identity of the simulated palm print image.

### Claim 6 (depends on 4)

The method according to claim 4, wherein
the one or more control vectors includes a random noise vector, and
the texture information includes one or more of light information, shadow information, and skin texture information.

### Claim 7 (depends on 1)

The method according to claim 1, wherein
the feature extraction is performed and the palm crease energy image is generated by a first generator,
the simulated palm print image is generated by a second generator, and
the first generator and the second generator are jointly trained based on a real palm print image.

### Claim 8 (depends on 7)

The method according to claim 7, further comprising:
jointly training the first generator and the second generator based on the real palm print image, wherein
the first generator uses a Bezier curve sample as an input, and outputs a palm crease energy image sample that includes a crease information sample,
the second generator uses a real palm crease energy image of the real palm print image as an input, and outputs a simulated palm print image sample that includes a texture information sample,
the real palm crease energy image is extracted from the real palm print image through a palm crease energy extractor,
the training the first generator and the second generator jointly is supervised based on the real palm crease energy image, and
the palm crease energy extractor is trained jointly with the first generator and the second generator.

### Claim 9 (depends on 8)

The method according to claim 8, further comprising:
extracting a set of real palm creases in the real palm print image through the palm crease energy extractor;
determining a line energy feature of each pixel in the set of real palm creases; and
obtaining the real palm crease energy image through a binarization process based on the set of real palm creases and the line energy feature of each pixel in the set of real palm creases.

### Claim 10 (depends on 8)

The method according to claim 8, wherein the first generator comprises a line feature enhancement block configured to enhance a line energy feature of a multi-channel feature map of the Bezier curve sample.

### Claim 11 (depends on 8)

The method according to claim 8, wherein
a joint loss function of the jointly training includes a first loss function related to the first generator and a second loss function related to the second generator,
the first loss function includes one or more of a contrastive loss configured to evaluate structural consistency between the palm crease energy image sample and the Bezier curve sample and an adversarial loss configured to evaluate similarity between the palm crease energy image sample and the real palm crease energy image, and
the second loss function includes one or more of a distribution control loss configured to evaluate similarity between a distribution of a control vector for generation of the simulated palm print image sample and a standard normal distribution, an identity consistency loss configured to evaluate similarity between the simulated palm crease energy image sample and the real palm crease energy image, a distortion loss configured to evaluate similarity between the simulated palm print image sample and the real palm image, and a discriminator loss configured to evaluate realism of the simulated palm print image sample.

### Claim 12 (depends on 1)

The method according to claim 1, wherein
the palm print curve template is pre-determined based on statistics obtained from a set of real palm creases of a human, and
the determining the plurality of control points includes:
determining regions based on the palm print curve template; and
obtaining the plurality of control points from the determined regions based on a sampling process.

### Claim 13

A palm print image generation apparatus, comprising:
processing circuitry configured to:
determine a plurality of control points based on a palm print curve template;
generate a set of Bezier curves based on the plurality of control points;
perform feature extraction on the set of Bezier curves, to obtain a feature map of the set of Bezier curves;
generate a palm crease energy image that includes a set of palm creases based on the feature map, the set of palm creases having a line distribution corresponding to a line distribution of the set of Bezier curves and having line stroke properties corresponding to line orientation energy of each pixel on the set of palm creases; and
generate a simulated palm print image having texture information based on the palm crease energy image.

### Claim 14 (depends on 13)

The apparatus according to claim 13, wherein the processing circuitry is configured to:
determine a line energy feature of each pixel in each per-channel feature map of the feature map that is a multi-channel feature map;
obtain an enhanced multi-channel feature map based on enhancement of the line energy feature of each pixel in each per-channel feature map of the feature map; and
generate the palm crease energy image based on the enhanced multi-channel feature map.

### Claim 15 (depends on 13)

The apparatus according to claim 13, wherein the processing circuitry is configured to:
generate the simulated palm print image based on the palm crease energy image and one or more control vectors configured to increase diversity among generated images.

### Claim 16 (depends on 13)

The apparatus according to claim 13, wherein
the feature extraction is performed and the palm crease energy image is generated by a first generator,
the simulated palm print image is generated by a second generator, and
the first generator and the second generator are jointly trained based on a real palm print image.

### Claim 17 (depends on 13)

The apparatus according to claim 13, wherein
the palm print curve template is pre-determined based on statistics obtained from a set of real palm creases of a human, and
the processing circuitry is configured to:
determine regions based on the palm print curve template; and
obtain the plurality of control points from the determined regions based on a sampling process.

### Claim 18

A non-transitory computer-readable storage medium storing instructions, which when executed by a processor, cause the processor to perform a palm print image generation method, the method comprising:
determining a plurality of control points based on a palm print curve template;
generating a set of Bezier curves based on the plurality of control points;
performing feature extraction on the set of Bezier curves, to obtain a feature map of the set of Bezier curves;
generating a palm crease energy image that includes a set of palm creases based on the feature map, the set of palm creases having a line distribution corresponding to a line distribution of the set of Bezier curves and having line stroke properties corresponding to line orientation energy of each pixel on the set of palm creases; and
generating a simulated palm print image having texture information based on the palm crease energy image.

### Claim 19 (depends on 18)

The non-transitory computer-readable storage medium according to claim 18, wherein the generating the palm crease energy image comprises:
determining a line energy feature of each pixel in each per-channel feature map of the feature map that is a multi-channel feature map;
obtaining an enhanced multi-channel feature map based on enhancing the line energy feature of each pixel in each per-channel feature map of the feature map; and
generating the palm crease energy image based on the enhanced multi-channel feature map.

### Claim 20 (depends on 18)

The non-transitory computer-readable storage medium according to claim 18, wherein the generating the simulated palm print image comprises:
generating the simulated palm print image based on the palm crease energy image and one or more control vectors configured to increase diversity among generated images.
