Research on the Measurement Method of the Detection Range of Vehicle Reversing Assisting System

Transcription

1 Research on the Measurement Method of the Detection Range of Vehicle Reversing Assisting System Bowei Zou and Xiaochuan Cui Abstract This paper introduces the measurement method on detection range of reversing assisting system, including reversing radar system and reversing vision system, which using laser distance measurementor and plannimeter. Fit the farthest positions that radar can detect smoothly in combination with the work principle of ultrasonic radar to realize visualization of detection range of reversing radar. A test bench was designed and established to simulate different installation size and location of the reversing assisting system of different types of vehicles. The deviation of test method was verified which can be controlled in a reasonable scope. Keywords Reversing assisting system Detection range Ultrasonic sensor Reversing vision system 1 Introduction Since the 21st century, intelligent vehicle technology is maturing progressively; it is an important part of Intelligent Transportation Systems (ITS). The main purpose of intelligent vehicle research and development is to reduce accident rates [1]. According to the domestic and international statistical data, the vehicle in reverse accident occurred frequently. A certain Chinese area traffic control department have done a micro statistics that there was 564 reverse collision accidents occurred within 20 days, almost 30 cases in one day, which not including the accidents solved privately. NHTSA statistics show that there are nearly 292 deaths accidents and injured accidents caused by vehicle reverse collision in one year. Behind these data is a potential cause, including but not limited to: vehicle performance and the extent of the driver s perception of reversing environment [2]. B. Zou (B) X. Cui China Automotive Technology and Research Center, Tianjin, China zoubowei@catarc.ac.cn X. Cui cuixiaochuan@catarc.ac.cn Springer International Publishing Switzerland 2017 V.E. Balas et al. (eds.), Information Technology and Intelligent Transportation Systems, Advances in Intelligent Systems and Computing 454, DOI / _15 55

2 56 B. Zou and X. Cui The NHTSA have extended requirements on the range of vehicle reversing radar and reversing image vision on the basis of FMVSS No.111, and accomplished large number of tests. ECE R46 and Japan and South Korea standards allow vehicles to install reverse assisting system, but is not in the scope of testing [3] Chinese standard, GB motor vehicle indirect vision device performance and installation requirements, also doesn t put forward performance requirements for reversing assisting system. This paper has proposed a measurement method for reversing assisting system in order to make up the merge of relative Chinese standards. 2 Working Principle of Reversing Radar The most commonly used sensors for reversing radar system are ultrasonic sensor. In addition, there are infrared and electromagnetic induction sensors. However, due to the weakness in anti-interference and limitation in usage, the latter ones are gradually eliminated by the market. The reversing assisting system is composed of ultrasonic sensors, control unit and warning unit. Ultrasonic sensors send signal and then receive the reflected signal when it encounter obstacles. The distance information is send to driver by warning unit after the calculation by control unit. The basic principle of ultrasonic ranging is called transit time method. Ultrasonic sensor signal strong electrical excitation, ultrasonic reflection waves are formed after the object being measured. Then the sensor calculates the distance to the objects detected according to the transmitting and receiving time. The transmitting time is t, and the transmitting velocity is c. Thus, the distance D, between sensor and object is: D = 1 2 c t (1) The acoustic energy along the axis of the sensor is the strongest and gradually fade out on both sides along the direction. Ultrasound beam width is defined as the angle where maximum acoustic energy down to the half. The narrower is the ultrasound beam width, the smaller is the detection range [4] Thus, the directivity of ultrasound is more obvious, either is the accuracy of detection range. There is a direct relationship between resonance frequency f and ultrasound beam width, and radiation area R. The relationship is expressed below: θ = 2arcsin 0.26c Rf (2) The commonly used frequency of ultrasonic sensor in reversing assisting system are 40, 48 and 58 khz [5]. The sensor with different frequency can measure different distance. The lower the frequency, the longer of the longitudinal length and the greater the vertical height can sensor measure. The OEMs have to consider the shape and size of vehicles to choose the appropriate reversing radar system to avoid the interference

3 Research on the Measurement Method 57 Fig. 1 Interference phenomenon phenomenon. As it shows in Fig. 1, the SUV and limousine were equipped with the same system. However, the installation height of limousine is too low, which result in a false alarming. Considering that the installation position of passenger car is relatively low and that the detection range of vertical direction is bigger, this paper only focuses on the research on measurement method of detection range in the horizontal direction. 3 Working Principle of Reversing Vision Comparing to the reversing radar system, reversing vision system provides more intuitive information. Most middle and high ranking passenger cars are equipped with both systems, which can improve the safety when reversing. Reversing vision system is composed of camera and display system. The camera is usually installed on the top of plate, pointing rearward. The display system is integrated to the central control panel. The information of vehicles tail environment is provided to driver directly. The working principle of reversing vision system can be explained by simple pinhole camera model. However, the model is an ideal linear model. Actually, reversing vision system uses exaggerated lens to gain wider field of vision, which brings in image distortion. Image distortion includes radial distortion and tangential distortion. There is no radial distortion in the center of the image plane, which is more and more serious when moving along the radial direction of image. Additionally, tangential distortion is due to the reasons for lens manufacturing flaws, causing that lens itself and the image plane are not parallel. In fact, in addition to the radial distortion and tangential distortion, there are many other distortion that can be neglected because of having no obvious influence [6].

4 58 B. Zou and X. Cui 4 Principle of Test According to characteristics of principle of reversing radar system and reverse vision system, this paper presents a test method to measure the detection range of both system respectively. This paper only did research on detection range in horizontal direction of reversing assisting system and put forward the following premises: The performance of reversing radar has the closet relations with installation position and pitch angle of radar. The ratio of basic level angle in horizontal and vertical direction is 2:1, about 120 :60. The detected object used in this paper can cover the vertical detection point of view. Thus, this paper did not carry out on vertical detection range measurement. This paper discusses the rear area of vehicle that cannot be observed from the rearview mirror, which is often located in the center of image, where with not serious distortion. Thus, the influence of lens distortion can be neglected. 4.1 Measurement Method on Detection Range of Reversing Radar System Make a car equipped with four-radar reversing assisting system as an example. Measure the installation height and angle of four radars before the test. In accordance with the premises, measure the detection range of each reversing radar. The specified test method as follow: (a) Horizontal detection distance measurement in radial direction. Power on only one radar sensor. Place the calibration object outside the area where radar can detect objects, and make sure that calibration object is in right front of radar and the radar emitting surface is parallel with the calibration object plane. Move the calibration object towards radar slowly until the warning system alarm constant audio cues. Then, stop moving the calibration object, and mark the position where it stops as point A. The distance from sensor to point A is defined as the radial detection distance (Fig. 2). (b) Horizontal lateral direction detection range measurement. Set sensor surface as center, choose one direction every (counter-clockwise in Fig. 3) five degree at a time, such as the direction 1, direction 2 and direction 3 direction X shown in Fig. 5. Use the same method mentioned above to measure detection distance in these directions and mark the positions (B, C and D in Fig. 3) respectively until the warning system no longer alarm. Similarly, the other side detection range can be measured with same method. (c) Connect all the points marked in each direction. The area surrounded by points and sensor is detection range of single radar. Repeat the measurement on other radars, then can acquire all radar detection ranges. Take the collection of all radar ranges, namely the entire vehicle detection range radar system.

5 Research on the Measurement Method 59 Fig. 2 Horizontal detection distance measurement in radial direction Fig. 3 Horizontal lateral direction detection range measurement 4.2 Measurement Method on Detection Range of Reversing Vision System The area directly behind the vehicle is significantly considered when measuring the detection range of reversing vision system. The steps of measurement are explained as following: (a) Determine the blind area directly behind vehicle according to the width of vehicle body, such as the shaded area shown in Fig. 8. (b) In the reversing image, determine four points, A, B, C and D, which are boundary points of blind area that could be seen from reversing image. (c) The area surrounded by four connected points is the reversing vision system can detected in the blind zone directly behind vehicle. Chinese mandatory standard GB Motor vehicles- Devices for indirect vision- Requirements of performance and installation puts forward regulations on interior view mirrors and main exterior view mirrors for passenger cars. The number of installation, mounting strength and field of view are specified in the standard.

6 60 B. Zou and X. Cui Fig. 4 Measurement on reversing vision system In the standard, The field of vision must be such that the driver can see at least a 20 m wide, flat, horizontal portion of the road centred on the vertical longitudinal median plane of the vehicle and extending from 60 m behind the driver s ocular points to the horizon (Fig. 4). The field of vision must be such that the driver can see at least a 4 m wide flat, horizontal portion of the road which is bounded by a plane parallel to the median longitudinal vertical plane passing through the outermost point of the vehicle on the passenger s side and which extends from 20 m behind the driver s ocular points to the horizon. In addition, the road must be visible to the driver over a width of 1 m, which is bounded by a plane parallel to the median longitudinal vertical plane and passing through the outermost point of the vehicle starting from a point 4 m behind the vertical plane passing through the driver s ocular points [7]. As can be seen from this standard, the requirements for the field of vision behind the vehicle areas specified area only after the driver s eye point mm extending to the ground, while the driver s eye point to mm area is not made demands. Define the shortest distance from vehicle tail to ground where driver can see from interior mirror as S1. D is the width of vehicle. The blind area directly behind vehicle that is not equipped with reversing assisting system S 0 is: S 0 = S1 D (3) Define S D is the area reversing assisting system can detect. The blind area directly behind vehicle that is equipped with reversing assisting system S S = S 0 S D (4) The most vulnerable group in revers collision accident is children according to the accidents statistics. This paper choose three different height of calibration object, including 50, 70 and 100 cm, to represent different posture and height of children.

7 Research on the Measurement Method 61 Measure the shortest distance from calibration object to vehicle tail when driver can see the object through interior mirror in the situation of different height of calibration object. The distance stands for the maximum length of blind zone in each situation. Finally, the proportion of detection range and blind zone can be measured and calculated. 5 Verification Test 5.1 Test Bench Design In this paper, test method was validated based on the test bench (see in Fig. 5). The test bench is constructed by aluminum structure and can simulate any dimension of vehicle tails equipped with different kinds of reversing assisting system. The test bench is composed of bench skeleton, brackets of sensor and laser emission system. The most significant is that the bracket of sensor can adjust the pitch angle and yaw angle of sensor (radar or camera) accurately. The main role of bench skeleton is to ensure the sensor can be adjusted in the direction of the three translational directions. For instance, in the vehicle coordinate system, the height from reversing sensors to ground and the distance between sensors in x and y directions can be altered. The function of bracket is to fix radar and camera, and to adjust the pitch angle and yaw angle of sensors (see in Fig. 6a). The horizontal and vertical compasses are used to indicate the installation pitch and yaw angle. The laser emission system has two functions, the first one is to display detection angle range of ultrasonic sensor, and the second one is to be the baseline which has different kinds of angle with normal line of sensor surface. The green lines in the Fig. 6b is the boundary of detection angle range of four ultrasonic radars. Fig. 5 Test bench

8 62 B. Zou and X. Cui Fig. 6 Sensor bracket (a)and laser emission system (b) 5.2 Test on Reversing Radar System This paper selected two passenger cars equipped with reversing radar system and reversing vision system as prototype. One is a midsize sedan, another is a SUV. Two different kinds of reversing assisting system products were selected for verification test. The two products both have reversing radar and reversing vision function. At the beginning of the test, the pressure of tire should be Tire pressure should be adjusted to the recommended cold tire pressure [8]. Then, accurately measure the installation dimension on prototype vehicle by using coordinate instrument. In accordance with the measurement data from coordinate instrument, adjust test bench to make the installation of sensors to be same with those on prototype vehicle.(see in Fig. 9a) The deviation can be controlled within 5 mm. Following the experimental method described earlier, the detection range points of each radar were marked on the ground (see in Fig. 6b). Connect the points of each radar on ground and add the boundary of vehicle width (see in Fig. 7). Then use planimeter to calculate the area of shadow part where radar system cannot detect. The results of tests are shown in Tables 1 and 2. This paper selected 3-D machine to detect the shortest distance that drive can observe from interior mirror. Using the principle of light reflection, in a position to determine whether the driver can observe the position of the driver by observing whether the light bulbs can be seen from interior mirror.(see in Fig. 9) The test results of vision minimum distance under the situation of three different calibration objects are indicated in Table 3. According to the minimum vision distance, Table 4 shows the proportion of blind area of different heights of calibration object (Fig. 8). At last, the test method of the radar detection range is verified, calculating the deviation of fitted boundary. Place the calibration objection the fitting boundary, if the warning system alarms, move the object outwards until alarm stop, recording the distance to the reversing radar. The difference between the distance from boundary to radar and the real distance from alarming position to radar is the deviation of the method. This paper chose several points on the boundary randomly to verify the deviation and the results were indicated in the Fig. 9. There is indication that the deviation of measurement can be controlled in 2 %.

9 Research on the Measurement Method 63 Fig. 7 The part of blind area radar cannot detect Fig. 8 Instrument used in test procedure. a Coordinate instrument. b D machine Fig. 9 Deviation of measurement method on radar detection range

10 64 B. Zou and X. Cui Table 1 Performance of reversing radar system on midsize sedan Midsize sedan Detect range unit: m 2 The proportion of detect range and blind area (%) Ultrasonic sensor A Ultrasonic sensor B Table 2 Performance of reversing radar system on SUV SUV Detect range unit: m 2 The proportion of detect range and blind area (%) Ultrasonic sensor A Ultrasonic sensor B Table 3 Minimum vision distance of different heights of calibration objects Vehicle type 50 cm 70 cm 100 cm Midsize sedan 13.3 m 9.3 m 2.47 m SUV 9.7 m 7.6 m 3.4 m Table 4 Proportion of blind area and vision field of different heights of calibration Vehicle type 50 cm 70 cm 100 cm Midsize sedan with A system Midsize sedan with B system SUV with A system SUV with B system Test on Reversing Vision System The camera is mounted on the vehicle s symmetrical vertical plane, the height of which is 947 and 1080 mm on midsize sedan and SUV respectively. Wide angle lens is generally used in the camera, the detection range is far and the lens distortion is large, but the detection of the near boundary is not coincident with the projection of the vehicle s tail profile. The test results are shown in Table 5, which mainly reflects the distance of the distance between the end of the detector and the distance to the surface projection, and the detection width nearest to the rear of the car. 6 Conclusion In this paper, a method for measuring the range of the assisting system is proposed, which is based on the reverse radar and reverse image system. In order to verify

11 Research on the Measurement Method 65 Table 5 The performance of reversing vision system on different vehicles Vehicle type Midsize sedan with A system Midsize sedan with A system SUV with A system SUV with B system Camera mounting 947 mm 947 mm 1080 mm 1080 mm height Distance between 310 mm 110 mm 225 mm 0mm rear of vehicle to nearest vision detection boundary Vision width nearest to the rear of the car 1690 mm 2450 mm 2045 mm 3130 mm the measurement method, the test carried out is based on the test bench, which can simulate installation size and location of the reversing assisting system of different types of vehicles. This paper used two types of vehicle, midsize sedan and SUV, for the tests. During the test, some conclusions can be concluded as follow: For the vehicle equipped with four-ultrasonic radar system, the detection range of the two sides of radar is less than those in the middle. For the vehicle equipped with reversing radar system, the blind area of vehicle is related to the distance of radar detection, but little to do with the detection angle range of radar. For the vehicle equipped with reversing vision system, the blind area of vehicle is related to the mounting height and mounting pitch angle. In the future, a measurement method based on vehicle should be put forward. To improve test efficiency, a more simplified measurement method should be developed due to the long test cycle, although the current test method can meet high accuracy. References 1. Hattori H (2000) Stereo for 2D visual navigation. Proc IEEE Intell Veh Symp, pp Department of transportation, National Highway Traffic Safety Administration, Federal Motor Vehicle Safety Standard; Rear Visibility; Final Rule (2014) 79(66), Department of transportation, National Highway Traffic Safety Administration. Low-speed Vehicles Phase-in Reporting Requirements, Federal Motor Vehicle Safety Standard (2010) pp Zhang H, Gao Y (2011) The research of ultrasonic ranging technique, instrument technology, , pp Feng Y, Wang Y (2011) Working principle of reversing radar. Operation and Maintenance, pp Ding H, Zou B (2013) Comparison of several lane marking line recognition methods. In: IEEE intelligent control and information processing, pp 53 58

12 66 B. Zou and X. Cui 7. GB Motor vehicles-devices for indirect vision-requirements of performance and installation (2013) 8. GB Limits of dimensions, axle load and masses for road vehicles (2004)