JPH03265100A - Image processing type traffic flow measuring instrument - Google Patents

Abstract

PURPOSE:To eliminate the influence caused by a shadow and to improve the discrimination of the kind of a car and the car number measuring accuracy by constituting the measuring instrument so that when a distance on a road surface of feature points of a car head and a side part of a vehicle is not varied from two-dimensional screen data subjected to image pickup from the upper part at a prescribed angle, it is detected as a part of the shadow. CONSTITUTION:An image pickup image is digitized from the upper part by a fixed camera 1, subjected to ternary conversion processing and a car head candidate is detected. When extracted two-dimensional data satisfies a reference value, a car head detecting part 24 decides it to be a vehicle, and a shadow discriminating part 25 tracks a timewise moving distance of the vehicle of every measuring period. The discriminating part 25 checks a variation of a distance between feature points of a car head part and a car side and in the case the distances on the road surface are equal irrespective of the lapse of time, its feature point is detected as a part of a vehicle shadow. A vehicle measuring part 2 sets this vehicle shadow as a noise, eliminates the part of the shadow from data of an image storage part 22 and recognizes only a car body. These traffic measured data are sent out to a central equipment from an output part 27. Accordingly, from an image measuring part 2, exact vehicle discriminated data and passing car number measured data which are not influenced by the shadow are obtained.

Description

[Detailed description of the invention]

[Industrial Application Field 1] The present invention relates to an image processing type traffic flow measurement device that detects vehicles through image processing using an industrial television (ITV) camera, etc. The present invention relates to a traffic flow measurement device that detects shadows and removes the effects of shadows.

[Conventional technology]

In this type of traffic flow measurement device, the shadows that appear around the vehicle may cause the size of the vehicle to be measured incorrectly, resulting in the erroneous detection of the vehicle type, or the number of vehicles being measured by being mistaken for another vehicle. This may cause incorrect detection. Therefore, it is necessary to accurately detect the shadows that appear around the vehicle and remove the effects of the shadows. The conventional method for detecting the shadow of a vehicle in this type of traffic flow measurement device is to detect the shadow by utilizing the fact that the shadow of the vehicle is darker than the road surface, and there is almost no change in brightness. is generally known (Hashimoto, Kumagai et al., “
“Development of Image Processing Traffic Flow Measurement System”, Sumitomo Electric Vol. 1
(Refer to No. 27, P59, September 1986). [Problems to be Solved by the Invention] However, in the conventional technology, as mentioned above, the conditions for determining a shadow are: ■ the brightness of the shadow screen is darker than the brightness of the road surface, and ■ there is almost no change in the brightness of the shadow. Since we only use There were issues that needed to be resolved, such as the shadows being detected as black vehicles. An object of the present invention is to solve the above-mentioned problems and provide an image processing type traffic flow measurement device that can accurately remove the influence of shadows and improve the accuracy of vehicle type discrimination and single-unit measurement.

[Means for Solving the Problems 1] In order to achieve the above object, the present invention provides an apparatus for measuring traffic flow by image processing image signals of traveling vehicles, which is installed at a predetermined height on the side of a road. , one imaging device that images a vehicle traveling on a road from above or from the front or rear at a predetermined fixed angle of view at a predetermined measurement cycle; and a plurality of two images taken by the imaging device at a time interval. calculation means for calculating the positions on the road surface of feature points of the vehicle head and side parts of the vehicle from the dimensional image data; A method in which it is determined whether the distance between the head and the feature point on the road surface does not change regardless of the passage of time, and if the distance does not change, the feature point is detected as part of the vehicle's shadow. The present invention is characterized by comprising a detection means. [Function 1] In the present invention, a plurality of two-dimensional screen data with time intervals are captured from above at a fixed angle of view by an imaging means (for example, an ITV camera) installed on the roadside at a predetermined height. is used to determine changes in the positional relationship between feature points on the front and side of the vehicle, and if the positional relationship does not change regardless of the passage of time, the feature point is detected as part of the shadow. In this way, in the present invention, the shadow of a vehicle is two-dimensional data that exists only on the road surface, and the shadow moves with the movement of the vehicle. Shadows are detected, so shadows are detected (
Shadows can be detected stably, and even if the shadow extends onto an adjacent lane, it will not be mistakenly detected as a black vehicle. Therefore, according to the present invention, the influence of shadows can be effectively removed, and the accuracy of vehicle type determination and the accuracy of vehicle number measurement can be improved. [Embodiment 1] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 shows the main circuit configuration of an image processing type traffic flow measuring device according to an embodiment of the present invention. In this figure, 1 is a camera unit (fixed camera) such as an ITV camera as an imaging means for capturing images of a vehicle running on a road, and 2 is a camera unit (fixed camera) that processes the output image signal of the camera unit 1 to generate images around the vehicle. This is an image measurement unit that detects shadows. The image measurement section 2 includes an A/D (analog/digital) conversion section 21. Image storage section 221 Image ternarization processing section 23. Vehicle head detection section 24. Shadow discrimination section 25. It is composed of a vehicle measuring section 26 and an output section 27. The A/D conversion section 21 digitizes the output image signal of the camera section 1. The image storage section 22 is composed of a RAM (random access memory), etc., and stores the digitized image signal as a luminance signal in units of imaging screens. . Image ternarization processing 23 performs ternarization (1, 0, -1) for each measurement sample point using the stored luminance signal. The vehicle head detection unit 24 detects the front part of the vehicle (hereinafter referred to as
(referred to as the head of the vehicle). The shadow determination unit 25 tracks the detected vehicle and determines the shadow of the vehicle from a plurality of two-dimensional image data as described below. The vehicle measuring section 26 is the shadow discriminating section 2
identify the vehicle using the vehicle shadow data obtained from 5;
Various types of traffic flow measurement data that are not affected by shadows are calculated, and the output unit 27 sends this measurement data to a remote solid device for traffic control. FIG. 2 shows an example of how the camera section 1 and image measurement section 2 shown in FIG. 1 are actually installed. Camera part l is multi-lane road 5
The image measurement unit 2 is fixed to the upper arm of a pillar (ball) 3 erected on the roadside 7 of , and the image measurement unit 2 is fixed at a predetermined position on the trunk of the pillar 3. The camera unit 1 images a predetermined range of a multi-lane road 5 from a certain height at a certain angle as a measurement area 4, and the image measurement unit 2 performs vehicle sensing processing from the image data captured by the camera unit 1. In this embodiment, a case where there are three lanes on each side will be considered as an example. In addition, 6 is in the center! 1 (median strip). Next, the operation of the embodiment of the present invention will be explained using FIGS. 3 to 7. FIG. 3 shows one screen imaged by the above-mentioned camera section l. This screen captures an image of a vehicle 10 (in this example, a large bus) traveling in the center lane from upstream to downstream, and the shadow 11 of this vehicle body extends to the adjacent lane. 4
The vehicle 10 is detected within the measurement area shown in , but this measurement area 4 is a plane parallel to the road surface, and its size (especially length) is such that it can detect the vehicle multiple times or more in each measurement cycle. . FIG. 4 schematically shows a vehicle candidate extraction screen in which the image data in FIG. 3 is ternarized (1, O, -1)L by the image ternarization processing unit 23. Characteristic points a to g in FIG. 4 are set by the vehicle head detection section 24. 5 and 6 are views of the state shown in FIG. 3 from above and from the side of the road, respectively, and show the current time t and previous times t' (that is, for a certain measurement period after time t). This shows the relative positional relationship between the vehicle and the shadow at the previous time. By using monocular stereoscopic viewing, which is one of the image processing techniques, even from the screen captured by a single camera as in this example,
The position of the object to be measured in three-dimensional space can be determined from the angle of view of the camera and the position of the road surface, which is the plane of the object to be measured. (2) The feature point g in FIG. 4 is a point on the optical axis 12 in FIG. 6 at time t°. (If the feature point g is part of the shadow, it matches the g+ point.) ②Now, consider three points (gZ+g's, g's) in FIG. 6 as the feature point g. These three points are located at different distances from the road surface, but from the camera unit 1, they all appear to exist on a point (force). Next, considering the state at time t, if the feature point g is a shadow (g'+ point in Figure 6), sunlight is parallel light, so a' - g' and a in Figure 5 -g's positional relationship on the road surface remains unchanged. Therefore, the distance between the distance points (O) and (R) from the vehicle head in FIG. 6 matches the distance between the points (A) and (1). ■On the other hand, when the feature point g is a part of a solid, that is, has a certain height from the road surface, for example, g's in Fig. 6, g
's), when the distance to the vehicle head is calculated at two times, time t' and time t, point (a) is obtained as shown in Figure 6.
Distance between points (A) and (A)Neither of the distances between points (A) and (T) match the distance between points (O) and (force). From the above, the positions of the vehicle head (points a, b) and feature point g on the road surface are calculated at each time and time t°. (Point (a) in Figure 6. (e) and point (1), (force)). Next, the distance between the vehicle head and the feature point is compared at each time, and if the distance is equal (if the distance is equal, the feature point g is detected as a point on the road surface, that is, as part of the shadow). In this way, the vehicle head portion a % in each cycle
From the relative positional relationship between b and each feature point Q % g, there are multiple points that exist on the road surface (points whose height from the road surface is "0") among feature points c - H. By recognizing this, a plane formed by a plurality of feature points existing on this road surface is detected as a shadow. Furthermore, the operation of the embodiment of the present invention will be explained in detail with reference to the flowchart of FIG. The output image signal of the camera section 1 is digitized by the A/D conversion section 21 and stored in the image storage section 22 periodically in units of screens at predetermined time intervals (step St). The image ternarization processing unit 23 calculates the road surface reference brightness (step S2), and calculates a threshold value for ternarization based on the road surface reference brightness (step S3). Here, the road surface reference brightness is calculated based on the input image data. This is a value obtained by changing the coefficient of exponential smoothing according to the magnitude of the difference value between the road surface reference brightness and the road surface reference brightness to follow the external environment. Furthermore, the image ternarization processing section 23 processes the luminance data of each measurement sample point read out from the image storage section 22 using the threshold value for ternarization described above (referred to as the ternarization threshold value). is ternarized (1, O. -1) (step S4). The vehicle head detection section 24 finds the luminance shift of the measurement sample points for each measurement line (road crossing direction) based on the ternarized data sent from the image ternarization processing section 23, and detects a vehicle head candidate. In other words, taking FIG. 4 as an example, each measurement line is scanned from the downstream direction at the bottom of the drawing to the upstream direction at the top of the drawing, and the length of the luminance shift in the measurement line direction is a predetermined constant length. If it is larger than , it is identified as the head portion (sides a, b) of the running vehicle (step S5). Next, the vehicle head detection unit 24 extracts the extracted two-dimensional data (
For the object), set 8 to g feature points as shown in Figure 4, calculate the position of each feature point on the measurement area surface, and calculate the brightness deviation width for each measurement line based on the vehicle length and vehicle length. Width judgment value (reference value)
Check whether you are satisfied with the following. If the determination value is satisfied, it is determined that the object is a vehicle. If the determination value is not satisfied, it is determined that the vehicle is not a vehicle and the step S described above is performed.
] (Step S6). When it is determined that the vehicle is a vehicle, the detection data of the vehicle head detection section 24 is passed to the shadow discrimination section 25, and the shadow discrimination section 25 calculates the temporal movement distance of the vehicle at preset time intervals, that is, at each measurement cycle. Track (Step SV). Next, the shadow discriminating unit 25 examines the change in the distance between the vehicle head and the feature points on the vehicle side (vehicle side) between time t and time t° (step S8), If the distances on the road surface are equal, the feature point is detected as part of the vehicle's shadow (step SIO). Also, time t
If the distance is shortened, it is determined that the vehicle is part of a single vehicle (step S9), and if the distance is longer, it is determined that the vehicle is part of a parallel vehicle (step S9).
). The result of determining the shadow of the vehicle is sent to the vehicle measuring section 26 in FIG. 1, and the process returns to step Sl to repeat the above-described process until the target vehicle leaves the measurement area 4 (step S12). The vehicle measuring section 26 uses the shadow of the vehicle detected by the shadow discriminating section 25 as noise, removes the shadow portion from the data in the image storage section 22, and recognizes only the vehicle body. This vehicle body recognition is performed, for example, by detecting the rise and fall of vehicle data within the measurement area.In other words, the measurement area is divided into an upstream area and a downstream area, and the rise of the vehicle data in the downstream area is recognized as a passing vehicle. We count the traffic volume, measure the vehicle length from the distance between the rise of vehicle data and the fall of vehicle data in the downstream area, and determine whether the vehicle is large or small. Calculate the speed (vehicle speed). These traffic flow measurement data are sent from the output unit 27 to a remote solid device. Therefore, the image measurement unit 2 can obtain accurate vehicle type discrimination data, passing vehicle number measurement data, etc. that are not affected by shadows. [Effect of the invention 1] As explained above, according to the present invention, the shadow of a vehicle is two-dimensional data that exists only on the road surface, and the shadow of the vehicle moves as the vehicle moves. multiple images taken from above at a fixed angle of view with a time interval.
From the dimensional screen data, if the positional relationship (distance) on the road surface between the feature points on the head of the vehicle and the side of the vehicle does not change regardless of the passage of time, it is detected as part of the shadow.
Shadows can be stably detected regardless of the color or shape of the vehicle in motion, improving the accuracy of vehicle type determination (large/small vehicle type discrimination). Since there is no false detection of a black vehicle, the accuracy of traffic volume measurement can be improved.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the circuit configuration of an embodiment of the present invention, Fig. 2 is a perspective view showing an example of installation of the device shown in Fig. 1, and Fig. 3 is a screen shot taken by the camera unit shown in Fig. 2. An explanatory diagram showing an example. Figure 4 is an explanatory diagram showing the screen state obtained by ternarizing the image data of one screen in Figure 3, Figure 5 is a plan view of the state in Figure 3 viewed from above the road, and Figure 6. 3 is a side view of the state shown in FIG. 3 viewed from the side of the road, and FIG. 7 is a flowchart showing the operation details of the embodiment of the present invention. ...Camera section, ...Image measurement section, ...Strut, ...Measurement area, ...Multi-lane road, 6.Central line, 7.Roadside, 1O..Vehicle, DESCRIPTION OF SYMBOLS 11... Shadow, 12... Optical axis, 21... A/D conversion part, 22... Image storage part, 23... Image ternary processing part, 24... Vehicle head detection part, 25...Shadow discrimination section, 26...Vehicle measurement section, 27...Output section.

Claims (1)

[Claims] 1) A device for measuring traffic flow by image processing image signals of traveling vehicles, which is installed at a predetermined height on the side of a road, and is installed at a predetermined height on the side of a road so that the vehicle traveling on the road can be viewed from above, in front of, or behind the vehicle. one imaging device that captures an image at a predetermined fixed angle of view at a predetermined measurement cycle; and a plurality of two-dimensional image data with time gaps captured by the imaging device to capture the head of the vehicle and the image of the vehicle. a calculation means for calculating the position of a side feature point on the road surface; and a calculation unit that calculates the position of the feature point on the road surface, based on the calculation result of the calculation means, when the feature point exists only on the road surface and the distance between the vehicle head and the feature point on the road surface. an image characterized by comprising a shadow detection means for determining whether or not the distance does not change regardless of the passage of time, and detecting the feature point as part of the shadow of the vehicle if the distance does not change. Processing type traffic flow measuring device.