JPH1114348A - Vehicle recognizing method using image system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate an error caused in a conventional image system by accurately discriminating whether it is a shadow formed under a vehicle or a general shadow, and recognizing an approaching vehicle by a discriminated result. SOLUTION: A portion where a gray level sharply varies is detected by successively comparing gray levels from one direction through a constant method on a signal of an image screen detected by a camera part. This method contains a shadow sensing stage to recognize only a shadow part, a forming stage to form a window, an operation stage to perform an operation on a dispersion value to brightness by a shadow set on an under surface of a formed window and a shadow discriminating stage to recognize the approach of a vehicle by recognizing a shadow of a road surface formed under the vehicle by comparing the arithmetically operated dispersion value and a preset reference value with each other.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for recognizing the presence of a vehicle using an image system.
An image system that accurately recognizes the shadow hanging between the vehicle's bumper and tire from the input camera image signal, and then determines the presence of an approaching vehicle through the recognized vehicle shadow. The present invention relates to a method for recognizing a vehicle.

[0002]

2. Description of the Related Art In general, an image system applied to an automobile is, in many cases, an apparatus mounted on an unmanned automobile and capable of controlling the traveling of the vehicle by itself. In many cases, the above-described unmanned vehicles identify the form and lane of the road from the front of the running vehicle, separate the road surface condition, and determine the approaching obstacle. , Monitor it slowly,
Automatically control the illuminating device, brake device and acceleration device,
As a result, the vehicle itself can run.

[0003]

However, the conventional video system recognizes the approaching opponent vehicle from the front of the running vehicle. Therefore, by identifying the shadow of the vehicle from the input video system, Although the vehicle can be recognized, an error has been generated in which the shadow of a building or a tree is mistaken for the shadow of the vehicle. Conventional video systems use a method for classifying approaching vehicles by identifying a shadow in contact with the road surface between a bumper and a tire in front of the approaching vehicle and classifying the approaching vehicle. In some cases, vehicles mistakenly perceived the shadows of buildings hanging above the building and the general shadows. In other words, the conventional video system has a problem that a vehicle mistakenly makes an error in the shade of a vehicle in a lane, a structure of a road, or a building or a tree hanging on the surface of the road.

According to the present invention, in order to solve the above-mentioned problem, gray levels are successively compared from a screen of an image in front of a vehicle input from a camera unit, so that shades are separated and subsequently classified. By calculating the variance value for the brightness from the shadow, it is accurately determined whether the shadow is formed at the lower part of the vehicle or a general shadow, and the approaching is performed according to the determined result. An object of the present invention is to provide a method for recognizing a vehicle using an image system capable of eliminating an error that has occurred in a conventional image system by recognizing the vehicle.

[0005]

SUMMARY OF THE INVENTION A portion where a gray level changes abruptly is detected by sequentially comparing the gray level of a signal of a video screen detected by a camera from one direction in a certain direction. And a shadow sensing step for recognizing only the shadow portion. In order to analyze the sensed shadow portion, the portion where the gray level changes rapidly is centered on the lower surface of the window. Forming a window, calculating a variance value for brightness from the shadow set on the lower surface of the formed window, and comparing the calculated variance value with a set reference value. And a step of recognizing the approach of the vehicle by recognizing the shadow of a road surface formed below the vehicle.

[0006] In addition, the step of identifying a shade includes a step of judging that the vehicle is shaded if the obtained variance is equal to or more than a reference value, and
And determining a general shade. Further, in the shadow identification step, a control signal is output according to the determined result to control a related device, and a gray level is compared in the input screen in which the gray level is compared after performing the control step. If there is a part for further comparing the levels, a routine step for allowing the gray levels to be compared at a slow speed is provided.

[0007]

BRIEF DESCRIPTION OF THE DRAWINGS FIG.
A method for recognizing the vehicle 3 (see FIG. 2) using the video system of the present invention will be described in detail based on an embodiment of the present invention. FIG. 1 is a block diagram showing a control state of the present invention and a general image system. The camera unit 10 is mounted on one side in front of a vehicle 3 and takes an image of an object 1. A main control unit 12 that outputs a control signal as a video signal, a steering control unit 14 that controls a steering device with a control signal of the main control unit 12, and a brake device that is controlled by a control signal of the main control unit 12. And an acceleration control unit 18 that controls an acceleration device based on a control signal from the main control unit 12.

As described above, the main control unit 12 of the video system configured, based on the video signal input through the camera unit 10, determines the object 1 in the traveling direction according to the present invention.
It is possible to separate the object 1 in more detail and determine the presence of the vehicle 3.

FIG. 2 shows a state in which a video signal input from the camera unit 10 is covered with a certain type of window 2 so that analysis can be easily performed. As shown in FIG. When the vehicle 3 is captured through the window 2, a line L <b> 1 in contact with the road surface can be searched for at the edge of the lower part of the object 1.

FIG. 3 is a diagram showing a state in which the image of a shadow and a road are sensed by an image system to which the present invention is applied. The main control unit 12 operates according to an image signal input from the camera unit 10. While sequentially comparing the gray levels from the bottom to the top, it is possible to search for all the shades 4 in which the brightness has changed abruptly.
By calculating a variance value for the brightness of the vehicle 3 and comparing the calculated value with a set value, it is possible to identify only the shadow 4 formed at the lower part of the vehicle 3.

For example, through a building or a vehicle 3 in a side lane,
It is possible for the formed shadow 4 to hang on the surface of the road ahead in which it travels. The shade 4 of the approaching vehicle 3 is
When the image system erroneously recognizes the shadow, the variance value for the brightness of the detected shadow 4 is compared with a set value, and based on the comparison result, the shadow 4 of the vehicle 3 is determined. Can be identified.

The relational expression and the theoretical explanation that can obtain the variance value for the brightness of the recognized shadow 4 are briefly described as follows. In the present invention, such a variance is referred to as a local variance in the present invention because the variance value for the brightness that can be obtained from the recognized shade 4 is analyzed by covering a window of the form. When the obtained shade 4 is subdivided into lattice stripes of a fixed size, a gray level for brightness can be obtained from each lattice stripe. For example, when the gray level that can be obtained from each subdivided grid pattern is in units of 8 bits, it is possible to classify up to 256 levels. In addition, the shade 4 subdivided and stylized by the above-described lattice fringe of a fixed size can obtain its variance (local variance) through the following relational expression.

[0013]

(Equation 1)

Μ is the average value for the gray level of the shade 4, N is the number of grids subdivided into shades, and Xi is the gray level (value of each function) of each subdivided grid.

In other words, the physical meaning of the local variance value that can be obtained through the above-mentioned relational expression indicates how large each lattice fringe is distributed from the average value. , Shadow 4 is very dark, its value (the local variance value described above) is close to 255, while shadow 4 is very light, it is close to zero.

Therefore, by comparing the obtained variance value (local variance value) with the set comparison value, and discriminating the magnitude of the value, the shadow 4 hanging at the lower part of the vehicle 3 is removed. It becomes possible to recognize, and as a result, it becomes possible to classify the approaching vehicle 3.

Further, like the illustrated histogram for local variance of FIG. 4, the distribution of the histogram is
When concentrated on one side, the variance value (local variance value) becomes small.

Contrary to the histogram shown in FIG. 4, in the case of the vehicle 3, the distribution of the histogram is not biased to one side as shown in the histogram shown in FIG. If the distribution is relatively uniform, the obtained variance (local variance) is large.

As described above, the image of the object 1 photographed by the camera unit 10 shows the degree of dispersion with respect to the brightness in the case of the single ship shade 4 and the case of the vehicle 4 shade 4. Are represented differently, and based on the comparison results, the main control unit 12 transmits a control signal when the vehicle 3 approaches by the steering control unit 14, the brake control unit 16, and the acceleration control unit 18. This makes it possible to automatically control the traveling of the vehicle 3 to an optimal state.

Therefore, in the video system according to the present invention, as shown in the control flowchart of the main control unit 12 shown in FIG. 6, the shade 4 of the vehicle 3 and the general shade 4 are transmitted to the camera unit 10. By discriminating and recognizing the input image screen and outputting the control signal, it is possible to prevent the object 1 from being erroneously recognized.

Further, when the video system according to the present invention is used in combination with the logic of a different video system, the recognition of the vehicle can be made more accurate and its reliability can be greatly improved.

A main control unit for inputting a video signal from the camera unit sequentially compares gray levels and sorts out portions where brightness changes rapidly, thereby separating shades. Through, by calculating the local variance value by a predetermined method, and comparing with the set reference value,
Accurate separation of vehicle shade and general shade is possible. Furthermore, by outputting a control signal based on the sorted result and automatically controlling the running of the vehicle, it is possible to prevent a mistake in recognition of an approaching vehicle and increase the reliability of the image system. Become.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a control state of a conventional video system.

FIG. 2 is a state diagram illustrating a state in which an input video signal is analyzed by applying a window of a certain format according to the present invention;

FIG. 3 is a diagram illustrating a state where an image of a shadow and a road is sensed by an image system to which the present invention is applied;

FIG. 4 is a histogram of local variance when classified into general shadows according to the present invention;

FIG. 5 is a histogram of local variance when classified according to the present invention behind a vehicle.

FIG. 6 is a control flowchart controlled by a main control unit to apply the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Target object 2 Window 3 Vehicle 4 Shadow 10 Camera section 12 Main control section 14 Steering control section 16 Brake control section 18 Acceleration control L1 Line where vehicle and road surface contact.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI H04N 5/225 G06F 15/70 410

Claims (3)

[Claims] An image screen signal detected by a camera is sequentially compared with a gray level in a certain direction from one direction to detect a portion where the gray level is rapidly changed. A shadow detecting step for recognizing only a portion, and a window for forming a window, wherein an abrupt change in gray level is centered on the lower surface of the window in order to analyze the detected shadow portion. A woo forming step, a calculating step of calculating a variance value for brightness from the shadow set on the lower surface of the formed window, and comparing the calculated variance value with a set reference value, A step of recognizing the approach of the vehicle by recognizing a shadow of a road surface formed under the vehicle. 2. The method of claim 2, wherein the step of determining whether the obtained variance is less than the reference value is performed when the obtained variance is equal to or greater than the reference value. 2. The method according to claim 1, further comprising the step of determining the vehicle as a general shadow. 3. The method according to claim 1, wherein the step of outputting a shadow includes a step of outputting a control signal according to a result of the determination, and a step of controlling a related device, and an input screen comparing gray levels after executing the control step. 2. The video system according to claim 1, further comprising a step of: when there is a part for further comparing the gray level, a step of allowing the gray level to be compared slowly. Vehicle recognition method using