KR101609819B1 - Apparatus for Inter-Vehicle Distance Estimation - Google Patents

Abstract

An apparatus for estimating an inter-vehicle distance includes a camera, an information processing device, a lane standard database, and an output device. The camera converts images of a preceding car and a lane to image data. The information processing device has a lane detection unit, a distance estimation unit, and a correction unit. The lane detection unit detects lane data by the image data being applied thereto from the camera for a plurality of frames. The distance estimation unit estimates a distance between the preceding car and a traveling vehicle by detecting a y-axis pixel position of the preceding car and comparing the y-axis pixel position to the lane data and pre-stored lane standard data. The correction unit corrects the estimated distance by collecting the distance estimated by the distance information generation unit for the plurality of frames. The lane standard data is stored in the lane standard database. The output device outputs the estimated distance generated by the information processing device.

Description

[0001] Apparatus for Inter-Vehicle Distance Estimation [

The present invention is to provide an inter-vehicle distance estimating apparatus. More particularly, to an inter-vehicle distance estimating apparatus which is simpler and more accurate in accuracy by using lane recognition.

Since the spread of automobiles, various driving assistance technologies have been developed along with the development of information and communication technologies. Driving assistance technologies include lane departure warning, sleep warning, cruise capable of cruising at speed, and distance warning from the preceding vehicle. However, most of the above technologies do not show enough accuracy or credibility to give their lives until they turn. In addition, the cost of application is high because it requires additional complicated additional equipment.

The technique of recognizing the inter-vehicle distance is being developed as an assistive technology for vehicle safety. In order to recognize the inter-vehicle distance, a method of comparing and analyzing images recognized by two cameras is generally used. In order to use images recognized by two cameras, there are various problems such as an increase in cost due to an increase in the number of cameras, and a free space for securing a separation distance between two cameras. This is not widely used because it is also related to the accuracy of the measured inter-vehicle distance.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an inter-vehicle distance estimating apparatus which is simpler and more accurate in accuracy by using lane recognition.

The present invention is a means for solving the above problems, wherein the inter-vehicle distance estimation apparatus includes a camera, an information processing apparatus, a lane standard database, and an output device. The camera converts the image of the leading vehicle and the lane into image data. The information processing apparatus includes a lane detecting unit that receives image data for a plurality of frames from the camera and detects lane data, and a lane detecting unit that detects a y-axis pixel position of the leading vehicle and compares the lane data and the stored lane standard data, And a correcting unit for correcting the estimated distance by collecting the distance estimated by the distance information generating unit during the plurality of frames. The lane specification database stores the lane specification data. The output device outputs the estimated distance generated by the information processing device.

In one embodiment, the information processing apparatus may further include a speedometer for measuring the speed of the driving vehicle, and the information processing apparatus may estimate the inter-vehicle distance using an average speed and a moving speed obtained for the plurality of feature points.

In one embodiment, the distance estimator may estimate the distance by comparing the number of lanes between the driving vehicle and the front vehicle and the number of adjacent lanes with the pre-stored inter-vehicle specification data.

As described above, according to the present invention, in the lane recognition and the headway distance estimation method, the manufacturing cost is reduced by using only one camera, and it is advantageous that the method can be used with only a narrow installation space. In addition, the accuracy of the headway distance estimation is improved by applying the method using the y-axis absolute position of the vehicle, the method using the feature points, and the speed as well as the method of simply recognizing the lane.

1 is a perspective view of an inter-vehicle distance estimator according to an embodiment of the present invention.
2 is an image showing the y-axis pixel position detection unit of Fig.
FIG. 3 is a graph for estimating the inter-vehicle distance using the y-axis pixel position detection unit of FIG.
4A to 4C are images showing the minutiae analyzing unit of Fig.
Fig. 5 is a plan view showing a lane detected by the lane detecting unit of Fig. 1; Fig.
Fig. 6 is an image showing a vehicle interval estimated using the lane detecting unit of Fig. 1; Fig.
FIG. 7 is a graph for estimating the inter-vehicle distance using the lane detecting unit of FIG.
Fig. 8 is an image showing a step of detecting a lane using the lane detecting unit of Fig. 1; Fig.
9A to 9C are images showing in detail the step of detecting the lane of FIG.
Fig. 10 is an image showing lanes detected using the images of Figs. 9a to 9c.
11 is a block diagram illustrating a method of processing lane detection information using the lane detecting portion of Fig.
12A to 12C are images showing a step of searching for a vehicle using the lane detecting portion of Fig.
13 is a flowchart showing a headway distance estimation method using the headway distance estimation apparatus shown in FIG.
14 is a flowchart showing the step of generating the inter-vehicle distance information in Fig.

Before describing in detail several embodiments of the invention, it will be appreciated that the application is not limited to the details of construction and arrangement of components set forth in the following detailed description or illustrated in the drawings. The invention may be embodied and carried out in other embodiments and carried out in various ways. It should also be noted that the device or element orientation (e.g., "front," "back," "up," "down," "top," "bottom, Expressions and predicates used herein for terms such as "left," " right, "" lateral," and the like are used merely to simplify the description of the present invention, Or that the element has to have a particular orientation.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings. Prior to this, terms and words used in the present specification and claims should not be construed as limited to ordinary or dictionary terms, and the inventor should appropriately interpret the concepts of the terms appropriately It should be construed in accordance with the meaning and concept consistent with the technical idea of the present invention based on the principle that it can be defined.

Therefore, the embodiments described in this specification and the configurations shown in the drawings are merely the most preferred embodiments of the present invention and do not represent all the technical ideas of the present invention. Therefore, It is to be understood that equivalents and modifications are possible.

1 is a perspective view of an inter-vehicle distance estimator according to an embodiment of the present invention.

1, an inter-vehicle distance estimating apparatus includes a camera 10, a speedometer 20, a lane standard database 40, an output device 50, and an information processing apparatus 100. [ At this time, the lane specification database 40 may be omitted. In another embodiment, it may further comprise a vehicle position measuring device, such as a GPS (not shown).

The camera 10 is disposed inside or outside the vehicle, and converts the front image into image data. For example, the camera 10 may include the same camera as the camera used for a general navigation system, a vehicle black box system, and the like.

The speed meter 20 generates a speed signal by measuring the speed of the vehicle in real time. For example, a speed signal can be received from a speedometer inside a traveling vehicle without including a separate speedometer. In another embodiment, the speed meter 20 may measure the absolute speed using a vehicle position measuring device such as a GPS or the like.

The lane specification database 40 includes lane specification data such as the inter-vehicle distance and the vehicle width for each road. For example, the lane specification database 40 may include lane specification data for each position.

The output device 50 outputs the inter-vehicle distance estimated by the information processing device 100.

The information processing apparatus 100 estimates an inter-vehicle distance from the next vehicle using the image data received from the camera 10, the speed signal received from the speedometer 20, and the lane specification data received from the lane specification database 40 .

The information processing apparatus 100 includes an image information processing unit 110, a distance information generation unit 120, a distance estimation unit 130, and a correction unit 140.

The image information processing unit 110 prepares the image data received from the camera 10 to remove noise. For example, when the luminance is too bright or dark, it can be compensated and a noise removing operation or the like can be performed. In addition, unnecessary left and right upper and lower images can be removed from the image data, so that only the image in the front direction can be extracted.

The distance information generation unit 120 generates distance information from the image data received from the image information processing unit 110. The distance information generation unit 120 may include a y-axis pixel position detection unit 122, a minutia analysis unit 124, a lane detection unit 126, and the like. The distance information generating unit 120 may include only one or two of the y-axis pixel position detecting unit 122, the feature point analyzing unit 124, and the lane detecting unit 126. [

The y-axis pixel position detection unit 122, the feature point analysis unit 124, and the lane detection unit 126 will be described later with reference to Figs. 2 to 12. Fig.

The distance estimating unit 130 estimates the distance between the driving vehicle and the front vehicle using the distance information received from the distance information generating unit 120.

The correction unit 140 corrects the distance estimated by the distance estimation unit 130 and re-applies the corrected inter-vehicle distance to the distance estimation unit 130. [

The output device 50 outputs the distance estimated by the information processing device 100. [

FIG. 2 is an image showing the y-axis pixel position detecting unit of FIG. 1, and FIG. 3 is a graph of estimating an inter-vehicle distance using the y-axis pixel position detecting unit of FIG.

1 to 3, the y-axis pixel position detection unit 122 detects the y-axis pixel position of the preceding vehicle in the image in the front direction received from the image information processing unit 110. [ For example, if the distance between the vehicle ahead and the vehicle is close, the y-axis pixel position of the leading vehicle in the frontal image is low (y1). When the distance between the vehicle ahead and the vehicle gradually increases, the position of the y-axis pixel gradually increases in the frontal direction image (y2, y3). If the distance between the vehicle ahead and the vehicle is far away, the position of the y-axis pixel is much higher in the frontal image (y4).

When the y-axis pixel positions (y1, y2, y3, y4) of the image of the preceding vehicle in the frontal direction image are detected using this, the headway distance between the y-axis pixel positions (y1, y2, y3, y4) d1, d2, d3, d4).

However, if the headway distance from the front vehicle is estimated using only the y-axis pixel position in the frontal direction image, an error may occur if the vertical height between the vehicle ahead and the vehicle is different. That is, when the driving vehicle passes the hill area, the y-axis pixel position of the leading vehicle in the uphill road is lowered, so that the estimated distance can be closer to the actual one. On the other hand, if the driving vehicle traverses the valley, the y-axis pixel position of the leading vehicle on the descending road becomes higher, and the estimated distance may be farther than the actual distance. Therefore, a technique for correcting the estimated distance using the minutia analysis unit 124 and the lane detection unit 126, which will be described later, is required.

4A to 4C are images showing the minutiae analyzing unit of Fig.

Referring to Figs. 1 to 3 and Figs. 4A to 4C, the inter-vehicle distance is estimated by using the characteristic points t1, t2, ..., tn of the image and the average speed of the vehicle.

For example, with respect to the positions y1, y2, ... yn sequentially from the position y1 of the forward vehicle to the position yn of the vehicle, the speeds v1, v2, ..., (d1, d2, ..., dn) at each pixel using the time (t1, t2, ..., tn)

[Formula 1]

 [Formula 2]

It is possible to obtain the inter-vehicle distances d3, d4, ..., dn at the remaining vehicle positions y3, y4, ..., yn in the same manner as in [Expression 1] and [Expression 2]. However, yn is the case where the distance between the driving vehicle and the front vehicle is zero, so it is not necessary to calculate the distance.

Fig. 5 is a plan view showing a lane detected by the lane detecting unit of Fig. 1; Fig.

Referring to FIGS. 1 to 3 and 5, the lane specification is defined for each road, and the vehicle-to-vehicle distance can be estimated using the width and the distance of the lane 5. For example, in the case of a highway or an accelerated road, the length of the lane 5 (white portion) is 10 m, the length between adjacent lanes 5 is 10 m, and the length of one lane set is 20 m (lane portion 10 m + The portion 10m between adjacent lanes). The length of the lane (5) is 3 m and the length between the adjacent lanes (5) is 5 m, and the length of one lane set is 8 m (lane portion 3 m + The part between the lanes is 5m). The length of the lane 5 is 5 m and the length between the adjacent lanes 5 is 5 m and the length of one lane set is 10 m (lane portion 3 m + between adjacent lanes 5) Part 5m). In the case of urban roads in the Busan area, the length of the lane 5 is 3 m and the length between the adjacent lanes 5 is 3 m, and the length of one lane set is 6 m (lane portion 3 m + 3m). In the case of the national road in the outlying area, the length of the lane 5 is 5 m, the length between the adjacent lanes 5 is 8 m, the length of one lane set is 13 m (lane portion 5 m + )to be. However, there are various lane sets such as 3m - 2m set, 3m - 3m set, 4m - 5m set, 4m - 7m set and 8m - 12m set in Jeju area.

The lane specification is stored in the lane specification database 40.

FIG. 6 is an image showing a roadway estimated using the lane detecting unit of FIG. 1, and FIG. 7 is a graph of estimating the roadway distance using the lane detecting unit of FIG.

1 to 3 and 5 to 7, the lane detecting unit 126 detects the image of the lane 5 and the y-axis pixel position of the next lane in the frontal direction image received from the image information processing unit 110. [ For example, the distance between the adjacent lanes (y2-y1) immediately preceding the driving vehicle, the lane distance thereafter (y3-y2), the adjacent lane distance thereafter (y4-y3) y4), it is possible to estimate the inter-vehicle distance (y5-y1) from the preceding vehicle.

For example, when the vehicle is traveling on a highway, the length of each lane and the distance between adjacent lanes are all 10 m, so the distance to the next vehicle is the sum of the distance α between the vehicle and the lane at 50 m.

Fig. 8 is an image showing a step of detecting a lane using the lane detecting unit of Fig. 1; Fig.

Referring to Figs. 1 to 3 and Figs. 5 to 8, the inter-vehicle distance includes a vertical position h1 of the vanishing point on the image data, a height h2 of the front portion of the vehicle (e.g., a monkey of the vehicle) h2) and the distance (w, a lane width between two points both sides of the lane is met), the number of runs (p _v), to meet the front part (h2) of the left lane and the moving vehicle left point (p _l), the right lane and the moving vehicle Can be obtained by using the right point P _r where the front portion h 2 of the first and second points meet.

9A to 9C are images showing in detail the step of detecting the lane of FIG.

Referring to Figs. 1 to 3 and Figs. 5 to 9C, both lanes are first detected to recognize lanes. In order to detect both lanes, only the pixels which are likely to be lanes in the image data (see Fig. 9A) are selected to make a separate image (see Fig. 9B). The size of a separate image, which is selected only for the pixels to be the lane, is equal to the size of the triangle formed by the vanishing point (p _v ), the left point (p _l ), and the right point (p _r ). At most two pixels are selected for each row of image data. The left edge is scanned in the left direction and the right direction is scanned in the right direction with the horizontal position of the vanishing point (p _v ) as the center, to select the edge pixels that meet first. In other embodiments, arbitrary bias values may be provided as margins at the start and end points of the scan to adjust the accuracy. If the lane is not found, the bias value is adjusted and a wider area can be scanned.

Fig. 10 is an image showing lanes detected using the images of Figs. 9a to 9c.

Referring to Figs. 1 to 3 and 5 to 10, straight lines are extracted from the image data obtained in Fig. 9C. Among the extracted straight lines, straight lines that can be candidates of the left lane and the right lane are selected. For example, the allowable angle of the left lane is 30 to 90 degrees, and the allowable angle of the right lane is -30 to 90 degrees. The straight lines disposed within the allowable angle among the straight lines that can become the candidates are respectively selected as the left lane and the right lane.

11 is a block diagram illustrating a method of processing lane detection information using the lane detecting portion of Fig.

1 to 3 and 5 to 11, a data set of the detected left lane, right lane, vanishing point (p _v ), left point (p ₁ ) and right point (p _r ) is accumulated. The accumulated data set can be obtained for each frame (frame 1, frame 2, ... frame n + 1). The accumulation of the data set continues until a predetermined threshold is reached. For example, the threshold may be n-1.

When a sufficient number of data sets exceeding the threshold value are collected, the lane position is estimated. The process of estimating the lane position and collecting and updating the data set is also repeated. Once a sufficient number of data sets over the threshold have been collected, the oldest data set is deleted as the data set of the new frame comes in each time. Therefore, the number of accumulated data sets is maintained by the threshold value (n-1).

To estimate the lane position, mean values of the vanishing point (p _v ), the left point (p _l ), and the right point (p _r ) are obtained. If the average values do not exceed the predetermined threshold values, it is determined that the vehicle is a correct lane. That is, between the vanishing point (p _v) means and a vanishing point in the current frame (p _v) the distance does not exceed the threshold value of the vanishing point between a left point (p _l) mean and left points (p _l) in the current frame of the if the distance does not exceed the threshold value of the left point, a distance between the right point (p _r) the average and the right point (p _r) of the current frame does not exceed the threshold of the right points, it is determined that the right lane.

On the other hand, when the average values exceed the predetermined threshold values, it is determined that the correct lane is not found. Further, when a change in the position and angle of the lane suddenly occurs, the lane departure (change) is determined.

12A to 12C are images showing a step of searching for a vehicle using the lane detecting portion of Fig.

Referring to Figs. 1 to 3 and Figs. 5 to 12C, among the inner regions (Fig. 12A) of the triangle formed by the vanishing point (p _v ), the left point (p _l ) and the right point (p _r ) the vanishing point (p _v) space (Fig. 12b) to that in the extracts. The extracted moving vehicle and the vanishing point (p _v) calculates the area histogram (Fig. 12c) of brightness values (shadow), for each line (row) (Fig. 12b) between. At this time, since the width in the horizontal direction decreases as the pixel approaches the vanishing point (p _v ), the ratio of the width of each row to the base w of the search area triangle is weighted.

The position of the next vehicle on the image data can be obtained using the histogram (Fig. 12C). In the present embodiment, the calculated histogram (Fig. 12C) is obtained every frame, and is recognized as the presence of the next vehicle only when it appears continuously for a predetermined number of times. In the case where the presence of the next vehicle is recognized only when the calculated histogram (FIG. 12C) continuously appears for a predetermined number of times, it is possible to prevent an error due to noise such as shade or stain on the road.

The inter-vehicle distance from the preceding vehicle is estimated using the position of the vehicle ahead, the inter-vehicle distance w, the length of the lane 5, and the like. The actual distance can be calculated using the recognized lane length using the lane length data stored in the lane specification database 40. [ In the present embodiment, a roughness distance according to the pixel height in the image data is calculated through a preprocessing process. For example, in the preprocessing process, the distance according to the height of the pixels in the image data may be calculated in advance, stored in the lane specification database 40, and the distance stored according to the pixel height of the detected next vehicle.

13 is a flowchart showing a headway distance estimation method using the headway distance estimation apparatus shown in FIG.

Referring to FIGS. 1 and 13, in order to estimate the inter-vehicle distance using the inter-vehicle distance estimating apparatus, image data is first generated using the camera 10. The generated image data is preprocessed using the image information processing unit 110 (S100). For example, adjustment of the luminance, elimination of errors caused by sunlight, streetlight, etc., and extraction of only a predetermined area in the frontal direction are performed.

Next, the inter-vehicle distance information is generated using the preprocessed image data (S200).

14 is a flowchart showing the step of generating the inter-vehicle distance information in Fig.

Referring to FIGS. 1, 13 and 14, in order to generate the inter-vehicle distance information, the y-axis pixel position is detected (S210), the feature point is analyzed (S220), and the lane can be detected (S230). The steps of detecting the y-axis pixel position (S210), analyzing the feature point (S220), and detecting the lane (S230) may be performed all three or only one or two of them may be performed. For example, only the step of detecting the lane (S230) may be performed. Steps S210 to S200 of analyzing the y-axis pixel position, S220 of analyzing the feature point, and step S230 of detecting the lane are the same as those shown in Figs. 2 to 12, and thus a duplicate description will be omitted.

Then, the inter-vehicle distance is estimated using the inter-vehicle distance information (S300). In this embodiment, the inter-vehicle distance is performed for a plurality of frames, and the inter-vehicle distances for the plurality of frames are corrected through the correction process (S400).

Finally, the inter-vehicle distance at which the error is eliminated is thrusted (S500).

According to the present invention, there is an advantage that the manufacturing cost is reduced by using only one camera in the lane recognition and the headway distance estimation method, and it is possible to use only a narrow installation space. In addition, the accuracy of the headway distance estimation is improved by applying the method using the y-axis absolute position of the vehicle, the method using the feature points, and the speed as well as the method of simply recognizing the lane.

Further, by correcting the estimated inter-vehicle distance due to the y-axis pixel position detecting unit using the feature point analyzing unit and the lane detecting unit, the accuracy of the estimated distance is improved even when the driving vehicle traverses the hill or the valley.

Furthermore, in estimating the inter-vehicle distance from the preceding vehicle, it is possible to recognize the presence of the next vehicle only when the distance is longer than a predetermined frame, thereby preventing an error caused by noise.

INDUSTRIAL APPLICABILITY The inter-vehicle distance estimating apparatus of the present invention is industrially applicable, which can be used in various fields such as a vehicle driving assist system, an unmanned operating system, a safe driving system of a locomotive or a subway.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. It is to be understood that various changes and modifications may be made without departing from the scope of the appended claims.

5: Lane 10: Camera
20: Speedometer 30: GPS
40: lane specification database 50: output device
100: information processing apparatus 110: image information processing section
120: distance information generation unit 122: y-axis pixel position detection unit
124: feature point analysis unit 126: lane detection unit
130: Distance estimation unit 140:

Claims (3)

A camera for converting the image of the leading vehicle and the lane into image data;
A lane detecting section for receiving image data from the camera for a plurality of frames and detecting lane data; and a lane detecting section for detecting a y-axis pixel position of the preceding vehicle and comparing the lane data and the stored lane standard data, An information processing apparatus comprising a distance estimator for estimating a distance and a compensator for collecting the distance estimated by the distance estimator during the plurality of frames and correcting the estimated distance;
A lane specification database in which the lane specification data is stored; And
And an output device for outputting the estimated distance generated by the information processing device
Wherein the distance estimating unit estimates the distance by comparing the number of lanes between the driving vehicle and the front vehicle and the number of adjacent lanes with the pre-stored inter-vehicle specification data.
delete delete

Images