# US 2015/0146922 A1 · Target detection device and target detection method

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

- Number: US 2015/0146922 A1
- Document ID: US20150146922A1
- Status: Published application
- Kind: A1
- Application number: 14/409,439
- Filed: 2013-06-28
- Published: 2015-05-28
- Earliest claimed: 2012-06-29
- Assignees: Secom Co., Ltd.
- Applicants: Secom Co., Ltd.
- Inventors: Takaharu Kurokawa; Takumi Munekata
- Web page: https://patentgrep.com/patent/US20150146922A1

## CPC

- G06V 10/75: Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V 10/255: {Detecting or recognising potential candidate objects based on visual cues, e.g. shapes}
- G06V 40/161: {Detection; Localisation; Normalisation}
- G06V 10/446: {using Haar-like filters, e.g. using integral image techniques}
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A target detection device that determines whether input data acquired from a data input module contains a detection target, the target detection device including: a multi-level data generation module for generating, from the input data, a plurality of data mutually different in an information level, the information level being a degree representing the detection target; an evaluation value calculation module for calculating, for each of the plurality of data, an evaluation value representing a degree of likelihood of the detection target; and a target determination module for determining that the input data contains the detection target when an increasing degree by which the evaluation value calculated for each of the plurality of data mutually different in the information level increases according to increase of the information level is equal to or more than a lower limit value of the increasing degree where the input data contains the detection target.

## Claims

### Claim 1

A target detection device that determines whether input data acquired from a data input module contains a detection target, the target detection device comprising:
a multi-level data generation module for generating, from the input data, a plurality of data mutually different in an information level, the information level being a degree representing the detection target;
an evaluation value calculation module for calculating, for each of the plurality of data, an evaluation value representing a degree of likelihood of the detection target; and
a target determination module for determining that the input data contains the detection target when an increasing degree by which the evaluation value calculated for each of the plurality of data mutually different in the information level increases according to increase of the information level is equal to or more than a lower limit value of the increasing degree where the input data contains the detection target.

### Claim 2 (depends on 1)

The target detection device according to claim 1, wherein the target determination module obtains, for each information level, a difference by subtracting the evaluation value of an information level lower by a predetermined number of levels than each information level from the evaluation value of each information level, and obtains any one of a total value and a maximum value of obtained differences, as the increasing degree.

### Claim 3 (depends on 1)

The target detection device according to claim 1, wherein the target determination module obtains a difference by subtracting, in ascending order of the information level, the evaluation value of an information level lower by a predetermined number of levels than each information level from the evaluation value of each information level, obtains, each time the difference is obtained, an accumulated value of the obtained difference as the increasing degree, determines whether the obtained increasing degree is equal to or more than the lower limit value, and determines that the input data contains the detection target when the obtained increasing degree is equal to or more than the lower limit value.

### Claim 4 (depends on 1)

The target detection device according to claim 1, wherein the target determination module obtains, as the increasing degree, any one of a difference obtained by subtracting a minimum value of the evaluation values of all the information levels from a maximum value of the evaluation values thereof and a ratio of the maximum value of the evaluation values of all the information levels to the minimum value of the evaluation values thereof.

### Claim 5 (depends on 1)

The target detection device according to claim 1, wherein the target determination module determines that the input data contains the detection target when the increasing degree is equal to or more than the lower limit value and a representative evaluation value which is a statistical representative value of the evaluation value calculated for each of the plurality of data mutually different in the information level is equal to or more than a predetermined value, whereas determines that the input data does not contain the detection target when the increasing degree is less than the lower limit value or the representative evaluation value is less than the predetermined value.

### Claim 6 (depends on 1)

The target detection device according to claim 1, wherein the information level is a degree of detail to which each of the plurality of data represents the detection target.

### Claim 7 (depends on 6)

The target detection device according to claim 6, wherein
the input data is image data; and
the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data mutually different in sharpness in which the sharpness is higher as the information level is higher, whereas the sharpness is lower as the information level is lower.

### Claim 8 (depends on 7)

The target detection device according to claim 7, wherein the multi-level data generation module generates, as the plurality of data mutually different in the sharpness, a plurality of data by applying an averaging filter which has a smaller filter size as the information level is higher and which has a larger filter size as the information level is lower, respectively.

### Claim 9 (depends on 6)

The target detection device according to claim 6, wherein the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data by superimposing, on the input data, a smaller amount of noise as the information level is higher and a larger amount of noise as the information level is lower, respectively.

### Claim 10 (depends on 6)

The target detection device according to claim 6, wherein
the input data is image data; and
the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data mutually different in image intelligibility in which the image intelligibility is higher as the information level is higher, whereas the image intelligibility is lower as the information level is lower.

### Claim 11 (depends on 10)

The target detection device according to claim 10, wherein the multi-level data generation module generates, as the plurality of data mutually different in the image intelligibility, a plurality of data by increasing a number of gradation levels of pixel values of the input data as the information level is higher and by reducing the number of gradation levels of the pixel values of the input data as the information level is lower.

### Claim 12 (depends on 1)

The target detection device according to claim 1, wherein the information level is a degree to which each of the plurality of data is suitable for feature representation of the detection target.

### Claim 13 (depends on 12)

The target detection device according to claim 12, wherein the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data by reducing a masking degree for a portion which represents the detection target in the input data as the information level is higher and by increasing the masking degree for the portion as the information level is lower.

### Claim 14 (depends on 12)

The target detection device according to claim 12, wherein the multi-level data generation module generates a feature parameter of the detection target, and generates, as the plurality of data mutually different in the information level, a plurality of data mutually different in information amount regarding the feature parameter such that the information amount represented by the feature parameter is larger as the information level is higher, whereas the information amount is smaller as the information level is lower.

### Claim 15 (depends on 14)

The target detection device according to claim 14, wherein the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data by multiplying respectively the feature parameter by a positive coefficient of 1 or less which is larger as the information level is higher and which is smaller as the information level is lower.

### Claim 16 (depends on 12)

The target detection device according to claim 12, wherein
the input data is image data; and
the multi-level data generation module generates, as the plurality of data mutually different in the information level, a plurality of data by geometrically converting the input data in a lesser degree as the information level is higher and in a greater degree as the information level is lower.

### Claim 17

A target detection method for determining whether input data acquired from a data input module contains a detection target in a target detection device, the target detection method comprising:
generating, from the input data, a plurality of data mutually different in an information level, the information level being a degree representing the detection target;
calculating, for each of the plurality of data, an evaluation value representing a degree of likelihood of the detection target; and
determining, by the target detection device, that the input data contains the detection target when an increasing degree by which the evaluation value calculated for each of the plurality of data mutually different in the information level increases according to increase of the information level is equal to or more than a lower limit value of the increasing degree where the input data contains the detection target.
