JPH0689339A - Body detecting method - Google Patents

Abstract

PURPOSE:To provide the body detecting method which requires no extra device except a camera, eliminates the need to prepare a reference image, and does not decrease in decision precision even if brightness varies. CONSTITUTION:An image input part 1 inputs an image photographed by the camera and outputs an image of a specific area predetermined in the image. A density distribution measurement part 2 inputs the image of the specific area, generates the density histogram of the image, and outputs the width W of a density distribution according to the density histogram. A detecting object presence/absence decision part 3 compares the width W of the density distribution with a threshold value sigma to decide whether or not the body to be detected is present. A decision result display part 4 displays the decision result of the detecting object presence/absence decision part 3.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object detecting method, and more particularly to an object detecting method for detecting whether or not an object to be detected exists in an image taken by an image pickup device such as a camera.

[0002]

2. Description of the Related Art Conventionally, as a method of detecting an object of this type, there are a method of detecting blocking of infrared rays and a laser beam, and a method of using a difference value between a reference image and an input image stored in advance.

[0003]

Among the above-mentioned conventional object detection methods, the former method has a drawback that another device is required in addition to a camera for taking an image,
In the latter method, the reference image must be prepared in advance, the reference image must be updated when the brightness changes, or an update error occurs, which increases the determination accuracy. There was a downside.

An object of the present invention is to provide an object detection method which does not require a special device other than a camera, does not need to prepare a reference image in advance, and does not reduce the determination accuracy even when the brightness changes. It is to be.

[0005]

A first object detecting method of the present invention comprises a step of creating a density histogram of a predetermined specific area of an image taken by an image pickup device, and a density distribution of the density distribution from the density histogram. The method includes a step of obtaining a spread amount of the density distribution, which represents spread, and a step of comparing the spread amount of the density distribution with a predetermined threshold value and determining that the detection target object is present when the spread amount is larger than the threshold value.

A second object detecting method of the present invention comprises the steps of creating a density histogram of a predetermined specific area of an image photographed by an image pickup device, and a density distribution showing the spread of the density distribution from the density histogram. , A step of obtaining a threshold value for detecting a detection target object from the density of the image, and a step of comparing the spread amount of the density distribution with the threshold value. Then, the step of determining is included.

[0007]

According to the present invention, the presence of the detection target object is determined by using the spread amount of the density distribution of the area where the detection target object is to be detected in the image, and the object detection is performed corresponding to the change in brightness. It is characterized in that the detection target object is detected by changing the threshold value for. Therefore, no special device other than the camera is required, and the determination accuracy does not deteriorate even when the brightness changes.

The amount of spread of the density distribution includes the maximum density value whose frequency of the density histogram is a predetermined first value or more and the minimum density value whose frequency is a predetermined second value or more. Although there are differences (called the width of the density distribution) and standard deviations of the density distribution, in the first embodiment below, the width of the density distribution will be used as the spread amount of the density distribution.
Further, the change in the brightness of the input image can be detected by using a predetermined specific area for object detection or by using a separately provided area. A description will be given using a predetermined specific area for detection. Further, the feature amount used to detect the change in the brightness of the input image has an average density value, a maximum density value, a median, etc., but the second embodiment will be described using the average density value.

[0009]

Embodiments of the present invention will now be described with reference to the drawings.

FIG. 1 is a block diagram of an apparatus for carrying out the object detecting method according to claim 1 of the present invention. The image input unit 1 inputs an image captured by a camera and outputs an image of a predetermined specific area from the image. The density distribution measuring unit 2 inputs an image of a specific area, creates a density histogram of the image, and outputs the width W of the density distribution from the density histogram. The detection target object presence / absence determination unit 3 compares the width W of the density distribution with a threshold value σ to determine the presence / absence of a detection target object. The determination result presentation unit 4 presents the determination result of the detection target object presence / absence determination unit 3.

FIG. 2 is a flow chart showing the processing of the density distribution measuring unit 2, and FIG. 3 is an explanatory view of the width W of the density distribution. The density distribution measuring unit 2 first receives an image of a predetermined specific area from the image input unit 1 (step 21), then creates a density histogram of the image (step 22), and finally, the density histogram. The difference between the maximum density value hmax of which the frequency is equal to or higher than a predetermined first value and the minimum density value hmin of the frequency is equal to or higher than a predetermined second value, that is, the width W = hmax-hmin of the density distribution. Is output to the detection target object presence / absence determining unit 3 (step 23). Here, the width W of the density distribution will be described with reference to FIG. 31 is the density histogram obtained in step 22, 32 is the maximum density value hmax whose frequency is greater than or equal to a predetermined first value,
33 is a minimum density value hmin whose frequency is a predetermined second value or more, and a difference hmax-hmi between 32 and 33.
n is the width W of the concentration distribution.

FIG. 4 is a flow chart showing the processing of the detection target object presence / absence determining unit 3. The detection target object presence / absence determining unit 3 first receives the width W of the density distribution obtained by the density distribution measuring unit 2 (step 41) and compares the threshold value σ held in advance with the width W of the density distribution (step S41). 42), if W is greater than σ, it is determined that there is a detection target object, and the determination result that “the detection target object is present” is output to the determination result presentation unit 4 (step 43). If W is not greater than σ, Judges that there is no detection target object, and outputs the judgment result of "no detection target object" to the judgment result presentation unit 4 (step 44). FIG. 5 is a diagram showing an example of the input image.

In the present embodiment, as shown in FIG. 5, an automobile passing on a road is set as a detection target object 52, and a predetermined specific area 51 is set as an area on the road. In the case of such an image, the width W of the density distribution of the specific area 51 is maximum when the vehicle is present on the specific area 51, and W is the minimum when the vehicle is only on the road surface, and the shadow of the oncoming vehicle or the like is in the specific area. 5
When multiplied by 1, takes an intermediate value. Therefore, the presence or absence of an automobile can be determined by checking the value of W.

As described above, according to this embodiment, it is possible to determine the presence or absence of the detection target object only by obtaining the width W of the density distribution and comparing it with the threshold value σ that is held in advance, without using the reference image. The object can be detected by a simple method.

FIG. 6 is a block diagram of an apparatus for realizing the object detecting method according to claim 2 of the present invention. The image input unit 61 inputs an image taken by a camera and outputs a predetermined specific image from the image. The density distribution measuring unit 62 inputs the specific image, creates a density histogram of the image, obtains the width W of the density distribution from the density histogram, and outputs the detection target object presence / absence determining unit 63 and the threshold setting unit 65. ,
Further, the average density value I of the predetermined area is obtained and output to the threshold setting unit 65. The threshold value setting unit 65 obtains the minimum width Wmin of the concentration distribution in the predetermined time T, and calculates W
The average density value Iwmin corresponding to min is output to the threshold table 66. The threshold table 66 shows the average density value Iw.
The threshold value σ corresponding to min is set as the detection target object presence / absence determining unit 63.
Output to. The detection target object presence / absence determining unit 63 compares the width W of the density distribution received from the concentration distribution measuring unit 62 with the threshold σ received from the threshold table 66 to determine whether or not the detection target object exists, and the determination result Is output to the determination result presentation unit 64. The determination result presentation unit 64 presents the determination result received from the detection target object presence / absence determination unit 63.

FIG. 7 is a flow chart showing the processing of the concentration distribution measuring unit 2. The density distribution measuring unit 62 first detects the image input unit 61.
The image is received from (step 71) and a density histogram of a predetermined area of the image is created (step 7).
2), the frequency of the density histogram is a predetermined first
Of the maximum density value hmax greater than or equal to the minimum density value and the minimum density value hmin greater than or equal to a predetermined second value, that is, the width W of the density distribution, h = hmax-hmin 63 and the threshold value setting unit 65 (step 73), and the average density value I of the predetermined area is obtained and output to the threshold value setting unit 65 (step 74).

FIG. 8 is a flow chart showing the processing of the threshold setting unit 65 and the threshold table 66. The threshold value setting unit 65 acquires the width W of the density distribution and the average density value I from the density distribution measuring unit 62 and stores them in the memory (step 8).
1) Next, if the width W of the obtained concentration distribution is smaller than the width of the concentration distribution held in the memory, replace the values W and I held in the memory with new ones respectively, and if the width W becomes smaller, the value of the memory Is left as it is (step 82). The minimum value Wmin obtained by repeating this operation for the time T is obtained, and the average density value Iwmin corresponding to the minimum value Wmin is output to the threshold value table 66 (step 8).
3) The threshold value σ corresponding to the average density value Iwmin is calculated in the threshold value table 66 and output to the target object presence / absence determining unit 63 (step 84).

FIG. 9 is a flow chart showing the processing of the detection target object presence / absence determining unit 63. The detection target object presence / absence determining unit 63 first determines whether the detection target object presence / absence determining unit 63 has a threshold value σ (step 91), and if not, determines whether the threshold value σ has been received from the threshold value table 66 (step 93). ), If not received, return to step 93, and if received, proceed to step 95. On the other hand, when it is determined in step 91 that the detection target object presence / absence determining unit 63 has the threshold value σ, the threshold value table 66 sets a new threshold value σ.
Is output (step 92). If a new threshold σ has not been output, the process proceeds to step 95. If a new threshold σ has been output, the detection target object presence / absence determining unit 63 currently holds. The threshold value σ is replaced with a new one output from the threshold value table 66, and the process proceeds to step 95. Step 9
In step 5, the width W of the concentration distribution is acquired, and it is checked whether or not the width W is larger than the threshold value σ (step 96). The output is performed (step 97), and if it is not large, it is determined that the detection target object 52 does not exist, and the determination result "no detection target object" is output (step 98).

In the present embodiment, as shown in FIG. 5, an automobile passing on the road is set as the detection target object 52, and a predetermined specific area 51 is set as the area on the road. In the case of such an image, the width W of the density distribution of the specific area 51 is maximum when the automobile is present on the specific area 51, and the width W of the density distribution is minimum when the vehicle is only on the road surface, and the width W of the oncoming vehicle or the like is reduced. When the shadow covers the specific area 51, it takes an intermediate value. Therefore, the presence or absence of an automobile can be determined by examining the value of the width W of the density distribution if the threshold value σ is set in advance according to the brightness of the image. As the brightness of the image, the average density value I of the specific area 51 when the specific area 51 is a road surface
Using, the correspondence table or the relational expression (threshold value table 66) between the average density value I and the threshold value σ is created in advance.

As described above, according to this embodiment, the brightness W of the input image is calculated by comparing the width W of the density distribution with the threshold value σ obtained using the width W of the density distribution without using the reference image. The presence or absence of the detection target object can be easily determined regardless of the change in

[0021]

As described above, the present invention has the following effects. (1) According to the first aspect of the present invention, the presence or absence of the detection target object is determined by using the spread amount of the density distribution of the area in the image where the detection target object is to be detected, without the need for a special device. , The detection target object can be easily detected. (2) According to the invention of claim 2, by changing the threshold value for object detection, the determination accuracy does not decrease even when the brightness of the input image changes.

[Brief description of drawings]

FIG. 1 is a block diagram of an apparatus for implementing an object detection method according to claim 1 of the present invention.

FIG. 2 is a flowchart showing a process of a concentration distribution measuring unit 2 in FIG.

FIG. 3 is an explanatory diagram of a width W of density distribution.

FIG. 4 is a flowchart showing a process of a detection target object presence / absence determining unit 3 in FIG. 1.

5 is a diagram showing an example of an image input to the image input unit 1. FIG.

FIG. 6 is a block diagram of an apparatus for carrying out the object detection method according to claim 2 of the present invention.

FIG. 7 is a flowchart showing the processing of the concentration distribution measuring unit 62 in FIG.

FIG. 8 is a threshold setting unit 65 and a threshold table 6 in FIG.
6 is a flowchart showing the processing of No. 6;

9 is a flowchart showing a process of a detection target object presence / absence determining unit 63 in FIG.

[Explanation of symbols]

 1 Image Input Unit 2 Density Distribution Measuring Unit 3 Detection Target Object Presence / Absence Determining Unit 4 Determination Result Presenting Unit 21-23 Step 31 Density Histogram 32 Maximum Density Value hmax 33 Minimum Density Value hmin 41-44 Step 51 Specific Area 52 Detection Target Object 61 Image input unit 62 Density distribution measurement unit 63 Detection target object presence / absence determination unit 64 Determination result presentation unit 65 Threshold setting unit 66 Threshold table 71-74, 81-84, 91-98 steps

Claims (2)

[Claims] 1. A step of creating a density histogram of a predetermined specific region of an image photographed by an image pickup device, and a step of obtaining a spread amount of the density distribution representing a spread of the density distribution from the density histogram, Compare the amount of spread of the concentration distribution with a predetermined threshold,
And a step of determining that an object to be detected is present when the object is larger than the threshold value. 2. A step of creating a density histogram of a predetermined specific area of an image captured by an image pickup device, and a step of obtaining a spread amount of the density distribution representing the spread of the density distribution from the density histogram. Object detection including a step of obtaining a threshold value for detecting a detection target object from the density of an image, and a step of comparing the spread amount of the density distribution with the threshold value and determining that a detection target object exists if the spread value is larger than the threshold value. Method.