JP2001028100A - Object extracting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately decide the possibility of a collision by extracting an object except an area whose horizontal length is equal to or greater than a prescribed length when an area rate is smaller than a prescribed rate. SOLUTION: A counter (j) is initialized to '1' (S31), and next, the area rate RATE of an object corresponding to the value of the counter (j) is calculated (S32). Whether or not the calculated area rate RATE is smaller than a prescribed rate RATETH is discriminated (S33), and in the case of the RATE < the RATETH, data whose horizontal length is equal to or greater than a prescribed length mTH is eliminated from the object (S34). Then, the counter (j) is increased (S35) and whether the value of the counter (j) surpasses the number (n) of objects is discriminated (S36). As a result, in the case of (j)>(n), the prescribed length mTH is made 1/2 (S37), an object label is updated (S38), and a divided object is extracted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object for extracting an object from an image obtained by, for example, an imaging device mounted on the vehicle, in order to determine the object that may collide with the vehicle. The present invention relates to an object extraction device.

[0002]

2. Description of the Related Art As a method of extracting an object from an image obtained by an image pickup device such as a television camera in order to determine the possibility of collision, for example, Japanese Patent Laid-Open No. 9-2592
No. 82 is known. This method extracts an optical flow on an image obtained by an imaging device, and determines an area having a flow other than the optical flow caused by traveling of the own vehicle as a moving obstacle area in the image, that is, a possibility of collision. And so on.

[0003]

However, the extraction of an object using such an optical flow requires a dedicated DSP or the like due to a large amount of calculation, which leads to an increase in cost. Therefore, it is conceivable that an infrared camera capable of detecting infrared rays can be used to easily extract a high-temperature target object such as an animal or a running vehicle. If, for example, a guardrail whose temperature is increased by sunlight is present in the background of the animal as the object, there is a problem that an image in which the animal and the guardrail overlap is extracted as the object.

The present invention has been made to solve this problem, and an object of the present invention is to provide an object extracting apparatus capable of easily and accurately extracting an object having a relatively high temperature.

[0005]

According to a first aspect of the present invention, there is provided an object extracting apparatus for extracting an object from an image obtained by an imaging means, wherein the imaging means detects infrared rays. It is possible to calculate the ratio of the area of the object on the image to the area of the image of the circumscribed rectangle including the object, and when the area ratio is smaller than a predetermined ratio, the length in the horizontal direction Is characterized in that the object is extracted except for an area longer than a predetermined length.

Here, the "predetermined ratio" is experimentally set to an optimal value, and the "predetermined length" is set to, for example, about 1/3 of the horizontal length of the image, or such a value is set. The same process is repeated by updating to a shorter value as an initial value. According to this configuration, the ratio of the area on the image of the target object to the area on the image of the circumscribed rectangle including the target object is calculated, and when the area ratio is smaller than the predetermined ratio, the length in the horizontal direction is predetermined. Since the target object is extracted by removing a region longer than the length, for example, even when a high-temperature target object and a guardrail overlap and are extracted, it is possible to separate and extract the target object. Become. Therefore, it is possible to accurately determine the possibility of collision by detecting the distance or the like for the object extracted in this manner.

[0007]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a configuration of a forward monitoring device for a vehicle including an object extracting device according to an embodiment of the present invention. The device includes two infrared cameras 1R and 1L capable of detecting far infrared rays, A yaw rate sensor 5 for detecting a yaw rate of the vehicle, a vehicle speed sensor 6 for detecting a traveling speed (vehicle speed) VCAR of the vehicle, a brake sensor 7 for detecting an operation amount of a brake, and these cameras 1R and 1L. An image processing unit 2 that extracts an object such as an animal in front of the vehicle based on the obtained image data and issues an alarm when there is a high possibility of a collision, a speaker 3 that issues an alarm by voice, and a camera 1R or 1L
And a head-up display (hereinafter, referred to as “HUD”) 4 for displaying an image obtained by the above-mentioned method and allowing the driver to recognize an object having a high possibility of collision.

As shown in FIG. 2, the cameras 1R and 1L are arranged in front of the vehicle 10 at positions substantially symmetrical with respect to the center axis of the vehicle 10 in the lateral direction.
The optical axes of L are fixed so that they are parallel to each other and their heights from the road surface are equal. Infrared camera 1R,
1L has the characteristic that the higher the temperature of the object, the higher the output signal level (the higher the luminance).

The image processing unit 2 includes an A / D conversion circuit that converts an input analog signal into a digital signal, an image memory that stores a digitized image signal, a CPU (Central Processing Unit) that performs various types of arithmetic processing, and a CPU that performs arithmetic operations. RAM used to store intermediate data (Ra
ROM (Read Only Memory) for storing programs, tables, maps, etc., executed by the CPU.
mory), an output circuit for outputting a drive signal for the speaker 3, a display signal for the HUD 4, and the like.
1L and the output signals of the sensors 5 to 7 are converted into digital signals and input to the CPU. As shown in FIG. 2, the HUD 4 is provided such that a screen 4a is displayed at a position in front of the driver in a front window of the vehicle 10.

FIG. 3 is a flowchart showing the procedure of processing in the image processing unit 2. First, the camera 1R, 1
The output signal of L is A / D converted and stored in the image memory (steps S11, S12, S13). The image stored in the image memory is a gray scale image including luminance information. FIG. 5A is a diagram for explaining a grayscale image obtained by the camera 1 </ b> R, and a hatched area is an area of an intermediate gradation (gray);
A region surrounded by a thick solid line is a region of the object having a high luminance level (at a high temperature) and displayed as white on the screen (hereinafter, referred to as a “high luminance region”).

In step S14 of FIG. 3, the image signal is binarized, that is, a luminance threshold value i determined as described later is set.
A process of setting an area brighter than TH to "1" (white) and setting a dark area to "0" (black) is performed. FIG. 5B shows an image obtained by binarizing the image shown in FIG. This figure shows that the hatched area is black, and the high-luminance area surrounded by the thick solid line is white.

FIG. 4 is a flowchart showing the binarization process in step S14 in detail. In step S21, as shown in FIG. 6A, the luminance value i is set on the horizontal axis, and the number of pixels taking the luminance value i is calculated. That is, a histogram having the frequency H [i] as the vertical axis is calculated. Normally, as shown in the figure, there is a background peak near the average luminance value of the background, and since the frequency of the object peak corresponding to the high-temperature object is low, it is difficult to identify the peak in the figure. . Therefore, in step S22, the histogram of the frequency H [i] shown in FIG. 9A is logarithmically converted as shown in FIG. 9B (G [i] = log ₁₀
H [i] +1) Calculate the histogram. This allows
It becomes easier to identify the object peak.

Next, a steepest descent method is applied to this histogram to search for a peak, and a luminance value i corresponding to the background peak is obtained.
PEAK1 and luminance value iPE corresponding to the object peak
AK2 is obtained (step S23). Then, a luminance threshold value iTH is calculated by the following equation (step S24). iTH = (iPEAK2-iPEAK1) × A + iPE
AK1

Here, A is a constant smaller than 1, and is set to, for example, 0.5. Next, an operation of converting the grayscale image data into the binary data using the luminance threshold value iTH is performed (step S25), and the present process ends. FIG.
In the processing of the luminance value corresponding to the background peak and the luminance value corresponding to the object peak using the histogram,
Since the luminance threshold value iTH is set between these luminance values, the object can be accurately detected as a high luminance area.

Returning to FIG. 3, in step S15, the binarized image data is converted into run-length data, and labeling for labeling each object is performed. FIG.
Is a diagram for explaining this, and in this diagram, white regions due to binarization are shown at the pixel level as lines L1 to L8. Each of the lines L1 to L8 has a width of one pixel in the y direction and is actually arranged without a gap in the y direction, but is shown apart for the sake of explanation.
Lines L1 to L8 each have two pixels in the x direction,
2 pixels, 3 pixels, 8 pixels, 7 pixels, 8 pixels, 8 pixels, 8
It has the length of a pixel. The run-length data indicates the lines L1 to L8 by using the coordinates of the start point of each line (the left end point of each line) and the length (the number of pixels) from the start point to the end point (the right end point of each line). It is shown. For example, the line L3 is composed of three pixels (x3, y5), (x4, y5) and (x5, y5), so that the run length data is (x3, y5, 3).

Next, as shown in FIG. 1B, a process of extracting the target object is performed by labeling the target object. That is, the lines L1 to L1 converted into run-length data
Lines L1 to L of L8 having portions overlapping in the y direction
3 is regarded as one object 1, the lines L4 to L8 are regarded as one object 2, and object labels 1 and 2 are added to the run-length data. By this processing, for example, FIG.
The high-luminance areas shown in (b) are grasped as objects 1 to 4, respectively.

In the following step S16, when the extracted object is integrated with the background or foreground, a process (FIG. 8) for separating the background or foreground is executed. Here, it is assumed that n objects have been extracted in step S15 of FIG. FIG. 9A illustrates an example in which the guardrail becomes hot due to sunlight and a high-temperature target is integrated with the guardrail and extracted as the target 1.
In such a case, the area SR of the circumscribed rectangle of the object 1
Ratio RATE of area SOBJ of object 1 to ECT
(= SOBJ / SRECT) is smaller than when only a high-temperature target is extracted. Thus, in the processing of FIG. 8, focusing on this point, the area corresponding to the guardrail is excluded from the target as the background (or foreground).

In step S31 of FIG. 8, the counter j is initialized to "1", and the area ratio RATE of the object j (the object corresponding to the value of the counter j) is calculated by the following equation (1) (step S31). S32). RATE = SOBJ (j) / SRECT (j) (1) Here, SOBJ (j) is the area of the object j on the image, and SRECT (j) is the area of the circumscribed rectangle of the object j. Area.

Next, the area ratio RATE is equal to the predetermined ratio RAT.
It is determined whether it is smaller than ETH (step S33),
If RATE ≧ RATETHH, the process immediately proceeds to step S35, while if RATE <RATETH, for example, as shown in FIG. Perform processing to delete from. Thus, the object recognized as the target object 1 is divided into three and recognized. The predetermined ratio RATETH is set to an optimum value based on various actually measured image data, and the initial value of the predetermined length mTH is, for example, the width of the horizontal direction of the image (the entire image captured by the camera 1R or 1L). Set to about 1/3.

In step S35, the counter j is incremented, and then it is determined whether or not the value of the counter j has exceeded the number n of objects (step S36). The process returns to step S32 while j ≦ n, and returns to step S32 when j> n.
At 37, the predetermined length mTH is halved, and then the object label is updated (step S38). As a result, the object divided in the process of step S34 is, for example, as shown in FIG.
As shown in (b), the objects are labeled as objects 1, 2, 3, and so on.

In the following step S39, a predetermined length mTH
Is determined to be equal to or less than a predetermined value M, and while mTH> M, the process returns to step S31 to repeat the same processing. When mTH ≦ M, this processing ends. Here, the predetermined value M is a normal horizontal width (horizontal width and length defined by the number of pixels) of the target object to be extracted at a preset distance.

As described above, according to the processing of FIG. 8, when the area ratio RATE is smaller than the predetermined ratio RATETH, a region whose horizontal length is equal to or longer than the predetermined length mTH is excluded from the target object. Even when a guardrail or the like and an object to be extracted are extracted as one object, it is possible to separate and extract them.

With respect to the extracted object, the distance from the host vehicle is calculated based on the images obtained from the two cameras 1R and 1L, and the possibility of collision is determined. By doing so, it is possible to accurately perform collision determination and the like. Further, as shown in FIG. 9B, the three divided objects 1, 2, and 3 can be integrally recognized based on the subsequent tracking data.
In the present embodiment, the image processing unit 2 constitutes an object extraction device.

[0024]

As described above in detail, according to the present invention, the ratio of the area of the object on the image to the area of the image of the circumscribed rectangle including the object is calculated, and the area ratio is more than the predetermined ratio. When it is small, the object whose length in the horizontal direction is equal to or greater than a predetermined length is removed and the object is extracted.For example, even when a high-temperature object and a guardrail overlap and are extracted, the object is extracted. Objects can be separated and extracted. Therefore, it is possible to accurately determine the possibility of collision by detecting the distance or the like for the object extracted in this manner.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a forward monitoring device for a vehicle including an object extracting device according to an embodiment of the present invention.

FIG. 2 is a diagram for explaining an arrangement of cameras.

FIG. 3 is a flowchart showing a procedure of object extraction.

FIG. 4 is a flowchart of a binarization process of FIG. 3;

FIG. 5 is a diagram for explaining a grayscale image and a binarized image.

FIG. 6 is a diagram illustrating a histogram of luminance values.

FIG. 7 is a diagram for explaining run-length data and labeling.

FIG. 8 is a flowchart of a separation process from a background.

FIG. 9 is a diagram for explaining the processing in FIG. 8;

[Explanation of symbols]

 1R, 1L infrared camera (imaging means) 2 Image processing unit (object extraction device)

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Masato Watanabe 1-4-1 Chuo, Wako-shi, Saitama Prefecture Inside Honda R & D Co., Ltd. (72) Inventor Hiroshi Hattori 1-4-1 Chuo, Wako-shi, Saitama F-term in Honda R & D Co., Ltd. (reference)

Claims (1)

[Claims] 1. An object extracting apparatus for extracting an object from an image obtained by an image pickup means, wherein the image pickup means is capable of detecting infrared rays, and has an area corresponding to an area of the circumscribed rectangle including the object on the image. Calculating a ratio of an area of the object on the image, and when the area ratio is smaller than a predetermined ratio, extracting the object except for an area whose horizontal length is equal to or longer than a predetermined length. Object extraction device.