# US 2015/0178588 A1 · Method and apparatus for recognizing object reflections

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

- Number: US 2015/0178588 A1
- Document ID: US20150178588A1
- Status: Published application
- Kind: A1
- Application number: 14/568,828
- Filed: 2014-12-12
- Published: 2015-06-25
- Earliest claimed: 2013-12-19
- Assignees: Robert Bosch GmbH
- Applicants: Robert Bosch GmbH
- Inventors: Michael Fischer
- Web page: https://patentgrep.com/patent/US20150178588A1

## CPC

- G06V 20/56: exterior to a vehicle by using sensors mounted on the vehicle
- G06V 10/60: relating to illumination properties, e.g. using a reflectance or lighting model
- G06V 20/582: {of traffic signs}
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A method for recognizing an object reflection in at least one image generated by a camera of a motor vehicle includes: recognizing at least one first object candidate in the image; identifying at least one first reflection indicator as a function of the first object candidate; evaluating the first reflection indicator to analyze whether the first object candidate is an object reflection; and outputting an evaluation signal as a function of the evaluation, the evaluation signal characterizing the first object candidate as at least one of an object reflection and an object.

## Claims

### Claim 1

A method for recognizing an object reflection in at least one image generated by a camera of a motor vehicle, comprising:
recognizing at least one first object candidate in the image;
identifying at least one first reflection indicator as a function of the first object candidate;
evaluating the first reflection indicator to determine whether the first object candidate is an object reflection; and
outputting an evaluation signal as a function of the evaluation, wherein the evaluation signal characterizes the first object candidate as at least one of an object reflection and an object.

### Claim 2 (depends on 1)

The method as recited in claim 1, further comprising:
recognizing at least one second object candidate in the image; and
identifying at least one second reflection indicator as a function of the second object candidate;
evaluating the at least one second reflection indicator to determine whether the second object candidate is an object reflection, wherein the evaluations of the first and second reflection indicators include comparing the first reflection indicator with the second reflection indicator; and
outputting the evaluation signal as a function of the evaluations of the first and second reflection indicators, wherein the evaluation signal additionally characterizes the second object candidate as at least one of an object reflection and an object.

### Claim 3 (depends on 2)

The method as recited in claim 2, wherein at least one of a geometric and color-related feature of at least one of the first and second object candidates is evaluated as a reflection indicator.

### Claim 4 (depends on 3)

The method as recited in claim 3, wherein a light characteristic of at least one of the first and second object candidates is evaluated as a reflection indicator.

### Claim 5 (depends on 1, 2, 3, 4)

The method as recited in claims 1 to 4, wherein a relative motion of at least one of the first and second object candidates relative to the camera is evaluated as a reflection indicator.

### Claim 6

A method for ascertaining an ambient condition in a sensing region of a camera of a motor vehicle, comprising:
recognizing object reflections in at least one image generated by a camera of a motor vehicle, by:
recognizing at least one first object candidate in the image;
identifying at least one first reflection indicator as a function of the first object candidate;
evaluating the first reflection indicator to determine whether the first object candidate is an object reflection; and
outputting an evaluation signal as a function of the evaluation, wherein the evaluation signal characterizes the first object candidate as at least one of an object reflection and an object;
ascertaining at least one reflection feature of the at least one object reflection;
ascertaining as a function of the reflection feature an ambient parameter describing an ambient condition; and
outputting the ambient parameter.

### Claim 7 (depends on 6)

The method as recited in claim 6, wherein a reflection location of the at least one object reflection is ascertained as the at least one reflection feature.

### Claim 8 (depends on 7)

The method as recited in claim 7, wherein an object location of an object associated with the at least one object reflection is ascertained, the ambient parameter being ascertained additionally as a function of the object location.

### Claim 9 (depends on 7)

The method as recited in claim 7, wherein a reflection dimension of the at least one object reflection is ascertained as an additional reflection feature.

### Claim 10 (depends on 7)

The method as recited in claim 7, wherein a reflection quality of the at least one object reflection is ascertained as an additional reflection feature.

### Claim 11

A non-transitory, computer-readable data storage medium storing a computer program having program codes which, when executed on a computer, perform a method for recognizing an object reflection in at least one image generated by a camera of a motor vehicle, the method comprising:
recognizing at least one first object candidate in the image;
identifying at least one first reflection indicator as a function of the first object candidate;
evaluating the first reflection indicator to determine whether the first object candidate is an object reflection; and
outputting an evaluation signal as a function of the evaluation, wherein the evaluation signal characterizes the first object candidate as at least one of an object reflection and an object.
