# US 2023/0410473 A1 · Methods for recognition of elements and element recognition device

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

- Number: US 2023/0410473 A1
- Document ID: US20230410473A1
- Status: Published application
- Kind: A1
- Application number: 18/333,969
- Filed: 2023-06-13
- Published: 2023-12-21
- Earliest claimed: 2022-06-14
- Assignees: Continental Autonomous Mobility Germany GmbH
- Applicants: Continental Autonomous Mobility Germany GmbH
- Inventors: Harinath Palavalli; Ashish Dhiman; Suma Sulibhavi; Vishant Garg
- Web page: https://patentgrep.com/patent/US20230410473A1

## CPC

- G06V 10/764: using classification, e.g. of video objects
- G06V 10/26: Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V 10/82: using neural networks
- G06V 20/49: {Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes}
- G06V 30/19173: {Classification techniques}
- G06V 30/1918: {Fusion techniques, i.e. combining data from various sources, e.g. sensor fusion}
- G06V 30/1985: {Syntactic analysis, e.g. using a grammatical approach (syntactic image representation G06V30/18171)}
- G06V 30/262: using context analysis, e.g. lexical, syntactic or semantic context
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A computer-implemented method for recognition of elements may include receiving input data that includes an observation of an object. The object may include a plurality of elements. The method may further include classifying each element of the plurality of elements more than once based on the observation of the object, resulting in a plurality of classifications for each element. The method may further include accumulating the plurality of classifications for each element, resulting in accumulated classifications for each element. The method may further include recognizing each element based on the respective accumulated classifications, resulting in a plurality of element recognitions. The method may further include recognizing the object based on the plurality of element recognitions.

## Claims

### Claim 1

A computer-implemented method for recognition of elements, the method comprising:
receiving input data comprising an observation of an object, the object comprising a plurality of elements;
classifying each element of the plurality of elements more than once based on the observation of the object, resulting in a plurality of classifications for each element;
accumulating the plurality of classifications for each element, resulting in accumulated classifications for each element;
recognizing each element based on the respective accumulated classifications, resulting in a plurality of element recognitions; and
recognizing the object based on the plurality of element recognitions.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein the input data comprises a plurality of datasets, wherein each dataset of the plurality of datasets comprises a respective observation of the object, and wherein classifying each element of the plurality of elements more than once comprises classifying each element based on the respective observation of the object from each dataset of the plurality of datasets.

### Claim 3 (depends on 2)

The computer-implemented method of claim 2, wherein the input data comprises a video, and wherein each dataset of the plurality of datasets comprises an image frame of the video.

### Claim 4 (depends on 2)

The computer-implemented method of claim 2, wherein each dataset of the plurality of datasets is obtained independently.

### Claim 5 (depends on 1)

The computer-implemented method of claim 1, wherein classifying each element of the plurality of elements more than once comprises classifying each element using a plurality of classifiers to determine a respective classification of the element by each classifier of the plurality of classifiers.

### Claim 6 (depends on 5)

The computer-implemented method of claim 5, wherein at least one classifier of the plurality of classifiers comprises a machine learning model.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, wherein the machine learning model comprises at least one of support vector machine, multi-layer perceptron and convolutional neural network.

### Claim 8 (depends on 1)

The computer-implemented method of claim 1, wherein recognizing each element based on the respective accumulated classifications comprises
sorting the accumulated classifications into a plurality of classification types,
counting a number of classifications in each classification type of the plurality of classification types, resulting in a count for each classification type, and
recognizing the element based on the classification type with a highest count.

### Claim 9 (depends on 1)

The computer-implemented method of claim 1, wherein recognizing each element based on the respective accumulated classifications comprises
sorting the accumulated classifications into a plurality of classification types,
counting a number of classifications in each classification type of the plurality of classification types, resulting in a count for each classification type, and
for each classification type of the plurality of classification types, computing a weighted score based on at least one of the count, confidence level, distance between the classifications, element size or track length, and
recognizing the element based on the classification type with a highest weighted score.

### Claim 10 (depends on 1)

The computer-implemented method of claim 1, wherein recognizing the object comprises aggregating more than one element recognition to form an element aggregate, and comparing the element aggregate to a plurality of predefined combinations of elements.

### Claim 11 (depends on 10)

The computer-implemented method of claim 10, wherein recognizing the object further comprises finding a closest match between the element aggregate and the plurality of predefined combinations of elements.

### Claim 12 (depends on 1)

The computer-implemented method of claim 1, further comprising:
recording the input data using a sensor,
wherein the sensor comprises a camera.

### Claim 13 (depends on 1)

An element recognition device comprising a processor configured to perform the computer-implemented method of claim 1.

### Claim 14 (depends on 13)

The element recognition device of claim 13, further comprising:
a mobility mechanism configured to move the element recognition device, also configured to adapt movement of the element recognition device based on the recognition of the object in the input data; and
a sensor configured to record the input data, wherein the sensor comprises a camera.

### Claim 15 (depends on 1)

A non-transitory computer-readable medium comprising instructions which, when executed by a computer, cause the computer to carry out the computer-implemented method of claim 1.
