Pedestrian detecting system

Patent metadata

Field	Value
Publication number	US-9727780-B2
Application number	US-201414586908-A
Country	US
Kind code	B2
Filing date	Dec 30, 2014
Priority date	Dec 30, 2014
Publication date	Aug 8, 2017
Grant date	Aug 8, 2017

Abstract

A pedestrian detecting system includes a depth capturing unit, an image capturing unit and a composite processing unit. The depth capturing unit is configured to detect and obtain spatial information of a target object. The image capturing unit is configured to capture an image of the target object and recognize the image, thereby obtaining image feature information of the target object. The composite processing unit is electrically connected to the depth capturing unit and the image capturing unit, wherein the composite processing unit is configured to receive the spatial information and the image feature information and to perform a scoring scheme to detect and determine if the target object is a pedestrian. The scoring scheme performs weighted scoring on a spatial confidence and an appearance confidence to obtain a composite scoring value to determine if the target object is the pedestrian.

First claim

What is claimed is:

1. A pedestrian detecting system, comprising:
a radar detecting and obtaining spatial information of a target object;
a camera capturing an image of the target object and recognizing the image, thereby obtaining image feature information of the target object; and
a computer electrically connected to the radar and the camera, wherein the computer receives the spatial information and the image feature information and performs a scoring scheme to detect and determine if the target object is a pedestrian;
wherein an appearance confidence is obtained through data transformation of the image feature information, and a spatial confidence is obtained through data transformation of the spatial information, and the scoring scheme performs weighted scoring on the spatial confidence and the appearance confidence to obtain a composite scoring value to determine if the target object is the pedestrian;
wherein the appearance confidence is calculated using the following equations:

P
appearance

⁡

(
B
)

=

1

1
+

e

-
α

×

tan

-
1

⁡

(

s
⁡

(

X
B

)

×

π
2

)

;

where s(X B ) represents a scoring value corresponding to a feature vector X in an image boundary region B of the target object;

the spatial confidence is calculated using the following equation:
P spatial ( B )= P associate ( c I ,c R *);
where P associate is a probability corresponding to each coupled position (C I , C R ), and C R * is a best matching that has a highest probability to each position C I ;

and the scoring scheme is calculated using the following equation:
P fusion ( B )=(1− w r )× P appearance ( B )+ w r ×P spatial ( B );

where Wr is a weighted ratio.

2. The pedestrian detecting system of claim 1 , further comprising an image processing apparatus, wherein the image processing apparatus obtains an intermediately-illuminated image region of the target object recognized by the camera, and retains the image feature information of the intermediately-illuminated image region, wherein an illuminance of the intermediately-illuminated image region is from 10 to 80000 lux.

3. The pedestrian detecting system of claim 2 , wherein the image processing apparatus obtains the intermediately-illuminated image region by using a Histogram of Oriented Gradient (HOG) and a Logarithm Weighted Pattern (LWP).

4. The pedestrian detecting system of claim 3 , wherein the operation of the Logarithm Weighted Pattern comprises capturing an image intensity variance of the target object, performing a logarithm operation and obtaining a portion of the image feature information.

5. The pedestrian detecting system of claim 1 , wherein the spatial information comprises a coordinate of the target object and a distance between the target object and a vehicle.

6. The pedestrian detecting system of claim 5 , wherein the scoring scheme uses a Gaussian function to obtain a matching probability between the image of the target object and the distance, thereby obtaining the spatial confidence.

7. A pedestrian detecting system, comprising:
a radar detecting and obtaining spatial information of a target object;
a camera capturing an image of the target object and recognizing the image, thereby obtaining image feature information of the target object;
an image processing apparatus obtaining an intermediately-illuminated image region of the target object recognized by the camera according to a relative luminance intensity of the image of the target object, and retaining the image feature information of the intermediately-illuminated image region, wherein an illuminance of the intermediately-illuminated image region is from 10 to 80000 lux; and
a computer electrically connected to the radar and the camera, wherein the computer receives the spatial information and the image feature information, and performs a scoring scheme to detect and determine if the target object is a pedestrian;
wherein an appearance confidence is obtained through data transformation of the image feature information, and a spatial confidence is obtained through data transformation of the spatial information, and the scoring scheme performs weighted scoring on the spatial confidence and the appearance confidence to obtain a composite scoring value to determine if the target object is the pedestrian;
wherein the appearance confidence is calculated using the following equations:

P
appearance

⁡

(
B
)

=

1

1
+

e

-
α

×

tan

-
1

⁡

(

s
⁡

(

X
B

)

×

π

Assignees

• Automotive Res & Testing Ct

Inventors

• Chang Kuo-Ching
• Liao Yu-Sheng
• Huang Pang-Ting
• Chan Yi-Ming
• Fu Li-Chen
• Hsiao Pei-Yung

Classifications

• G06V10/60

  relating to illumination properties, e.g. using a reflectance or lighting model · CPC title

• G06V10/50

  by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis · CPC title

• G06V10/467

  Encoded features or binary features, e.g. local binary patterns [LBP] · CPC title

• G06K9/00805

  Physics · mapped topic

• G01B11/14

  for measuring distance or clearance between spaced objects or spaced apertures (G01B11/26 takes precedence; rangefinders G01C3/00) · CPC title