# US 2021/0334515 A1 · Deep learning based image analysis method, system and mobile terminal

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

- Number: US 2021/0334515 A1
- Document ID: US20210334515A1
- Status: Published application
- Kind: A1
- Application number: 16/619,924
- Filed: 2019-03-27
- Published: 2021-10-28
- Earliest claimed: 2018-11-14
- Assignees: TUAT Co.,Ltd.
- Applicants: TUAT Co.,Ltd.
- Inventors: Soo Won CHO
- Web page: https://patentgrep.com/patent/US20210334515A1

## CPC

- G06V 40/166: {using acquisition arrangements}
- H04M 1/72403: with means for local support of applications that increase the functionality
- G06F 3/04883: for inputting data by handwriting, e.g. gesture or text
- G06N 3/02: Neural networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06V 10/764: using classification, e.g. of video objects
- G06V 10/82: using neural networks
- H04N 21/23418: {involving operations for analysing video streams, e.g. detecting features or characteristics (television picture signal circuitry for scene change detection H04N5/147; filtering for image enhancement G06T5/00; methods or arrangements for recognising scenes G06V20/00; arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59)}
- H04N 21/251: {Learning process for intelligent management, e.g. learning user preferences for recommending movies (details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08)}
- H04N 21/6582: {Data stored in the client, e.g. viewing habits, hardware capabilities, credit card number (arrangements where receivers interact with the broadcast H04H20/38)}
- H04N 21/8153: {comprising still images, e.g. texture, background image}
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- H04M 2250/52: including functional features of a camera
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- H04M: TELEPHONIC COMMUNICATION (circuits for controlling other apparatus via a telephone cable and not involving telephone switching apparatus G08)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

An image analysis server based on deep learning is configured to include an image analysis mobile terminal for photographing an image and transmitting the photographed image; and an image analysis server for receiving the photographed image, generating analysis result values through neural networks corresponding to multiple functions, incorporating priorities according to weight into probability values of the generated analysis result values, and providing result values to the image analysis mobile terminal.

## Claims

### Claim 1

An image analysis server based on deep learning, comprising:
an image analysis mobile terminal for photographing an image and transmitting the photographed image; and
an image analysis server for receiving the photographed image, generating analysis result values through neural networks corresponding to multiple functions, incorporating priorities according to weight into probability values of the generated analysis result values, and providing result values to the image analysis mobile terminal.

### Claim 2 (depends on 1)

The image analysis server of claim 1, wherein the image analysis server receives the photographed image and generates the analysis result values through an image description neural network, a facial recognition neural network and a text recognition neural network.

### Claim 3 (depends on 2)

The image analysis server of claim 2, wherein the image analysis mobile terminal displays an analysis result value having a highest priority among the analysis result values generated through the image description neural network, the facial recognition neural network and the text recognition neural network.

### Claim 4 (depends on 3)

The image analysis server of claim 3, wherein the image analysis mobile terminal displays the analysis result values in order of higher priority in response to a received swipe command.

### Claim 5

An image analysis method based on deep learning, comprising:
a first step of photographing, by an image analysis mobile terminal, an image and transmitting the photographed image;
a second step of receiving, by an image analysis server, the photographed image and generating analysis result values through neural networks corresponding to multiple functions; and
a third step of incorporating, by the image analysis server, priorities according to weight into probability values of the generated analysis result values and providing result values to the image analysis mobile terminal.

### Claim 6 (depends on 5)

The image analysis method of claim 5, wherein the second step comprises receiving, by the image analysis server, the photographed image and generating the analysis result values through an image description neural network, a facial recognition neural network and a text recognition neural network.

### Claim 7 (depends on 6)

The image analysis method of claim 6, further comprising a fourth step of displaying, by the image analysis mobile terminal, an analysis result value having a highest priority among the analysis result values generated through the image description neural network, the facial recognition neural network and the text recognition neural network.

### Claim 8 (depends on 7)

The image analysis method of claim 7, further comprising a fifth step of displaying, by the image analysis mobile terminal, the analysis result values in order of higher priority in response to a received swipe command.

### Claim 9

An image analysis mobile terminal based on deep learning, comprising:
an image photographing unit for photographing an image;
a communication unit for transmitting the photographed image to an image analysis server, and receiving, from the image analysis server, results in which priorities according to weight have been incorporated into probability values of analysis result values generated through neural networks corresponding to multiple functions;
a touch display unit for receiving display and control commands for information; and
a controller for controlling the touch display unit to display an analysis result value having a highest priority among the analysis result values.

### Claim 10 (depends on 9)

The image analysis mobile terminal of claim 9, wherein the controller controls the communication unit to display, through a touch display unit, the analysis result value having the highest priority among the analysis result values, which have been received from the image analysis server and into which priorities have been incorporated through an image description neural network, a facial recognition neural network and a text recognition neural network.

### Claim 11 (depends on 9)

The image analysis mobile terminal of claim 9, wherein the controller displays the analysis result values in order of higher priority in response to a swipe command received through the touch display unit.
