Object detection system and method

Abstract

An object detection system and method. The object detection system includes a transmitter, a receiver, and a processing circuit; the processing circuit is configured to: control the transmitter to emit a plurality of detection signals in different time frames toward a main beam direction with a predetermined field pattern; control the receiver to receive a plurality of reflected signals; calculate a plurality of corresponding received powers, a plurality of distances, and a plurality of velocities; perform a grouping process on the distances and the velocities to find out the received power, the distance, and the velocity corresponding to a main target; perform a correlation process to track the distance and the received power of the main target in different time frames; and calculate the power variation trend and the distance variation trend of the main target, and determine whether a pre-warning event will occur according to the relationship between the power variation trend and the distance variation trend. The object detection system and method of the present application allow the trajectory of the object to be inferred without the need for additional modification of hardware, and are suitable for situations where it is not necessary to know the angle of the object relative to the transmitter with very high precision and for multiple target objects.

Description

Technical Field

[0001] This invention relates to an object detection system and method, and more particularly to an object detection system and method for determining whether an early warning event will occur. Background Technology

[0002] A radar system has a radio frequency (RF) module for controlling the waveform, a transmitter (TX), and a receiver (RX). The signal emitted by the radar system will strike an object and be reflected back to the receiver, which processes the received signal to determine the range and speed of the object that the signal struck and reflected back.

[0003] In existing radar systems, at least two such... Figure 1 The receiver antenna shown is used to determine either the azimuth or the elevation angle, and at least three antennas are required to determine both the azimuth and elevation angles simultaneously.

[0004] Furthermore, radar systems have radio frequency (RF) front-end circuitry to control the transmitter and receiver. For the multiple antennas mentioned above, each antenna requires an analog-to-digital converter circuit to convert analog signals into digital signals for further processing by the back-end processing circuitry. Therefore, the addition of RF front-end circuitry will increase the cost and complexity of the system.

[0005] Furthermore, when the aforementioned radar system is used as an automotive radar system, a more complex radar system must be used in conjunction with a camera in order to further predict the trajectory of objects, thus resulting in higher cost and complexity.

[0006] Therefore, there is a need to provide an object detection system and method to solve the above problems. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide an object detection system and method for determining whether an early warning event will occur, in order to address the shortcomings of the prior art.

[0008] To address the aforementioned technical problems, one technical solution adopted by the present invention is to provide an object detection system, comprising a transmitter, a receiver, and a processing circuit. The processing circuit is connected to the transmitter and the receiver and is configured to: control the transmitter to transmit multiple detection signals in a predetermined field pattern toward a main beam direction at different time frames, wherein the main beam direction corresponds to a main beam in the predetermined field pattern generated by the transmitter through beamforming; control the receiver to receive multiple reflected signals generated by the reflection of the detection signals; calculate multiple received powers, multiple distances, and multiple velocities based on the reflected signals; perform a clustering process on the distances and velocities to identify the received powers, distances, and velocities corresponding to at least one primary target; perform an association process to track the distances and received powers of the at least one primary target in different time frames; and calculate a power change trend and a distance change trend of the at least one primary target, and determine whether a warning event will occur based on the relationship between the power change trend and the distance change trend. Specifically, in response to the distance change trend indicating that the at least one main target is approaching, and the power change trend indicating that the received power corresponding to the at least one main target is increasing, it is determined that an early warning event will occur with the at least one main target.

[0009] To address the aforementioned technical problems, another technical solution adopted by the present invention is to provide an object detection method. This detection method is applicable to a detection system, which includes a transmitter, a receiver, and a processing circuit. The detection method includes configuring the processing circuit to: control the transmitter to transmit multiple detection signals in a predetermined field pattern toward a main beam direction at different time frames, wherein the main beam direction corresponds to a main beam in the predetermined field pattern generated by the transmitter through beamforming; control the receiver to receive multiple reflected signals generated by the reflection of the detection signals; calculate multiple received powers, multiple distances, and multiple velocities based on the reflected signals; perform a clustering process on the distances and velocities to find the received powers, distances, and velocities corresponding to at least one main target; perform an association process to track the distances and received powers of the at least one main target in different time frames; and calculate a power change trend and a distance change trend of the at least one main target, and determine whether a warning event will occur based on the relationship between the power change trend and the distance change trend. Specifically, in response to the distance change trend indicating that the at least one main target is approaching, and the power change trend indicating that the received power corresponding to the at least one main target is increasing, it is determined that an early warning event will occur with the at least one main target.

[0010] One of the advantages of the present invention is that the object detection system and object detection method provided by the present invention will allow the trajectory of the object to be inferred without additional hardware modifications, and are applicable to situations where it is not necessary to know the angle of the object relative to the transmitter with very high precision, as well as situations involving multiple target objects.

[0011] This invention does not require highly precise determination of the object's angle. Instead, it determines the relationship between distance and power change trends by calculating information such as the distance of the reflected signal, the distance between the object and the receiver, and the received power. When it is only necessary to guess whether an object constitutes an obstacle, the required number of antennas is smaller, resulting in fewer corresponding RF front-end circuits and analog-to-digital converters. Therefore, the object detection system and method of this invention are superior to existing systems in terms of cost and complexity.

[0012] To further understand the features and technical content of the present invention, please refer to the following detailed description and accompanying drawings. However, the drawings provided are for reference and illustration only and are not intended to limit the present invention. Attached Figure Description

[0013] Figure 1 This is a functional block diagram illustrating an object detection system according to an embodiment of the present invention.

[0014] Figure 2 This is a flowchart illustrating a detection process according to an embodiment of the present invention.

[0015] Figure 3 This is a flowchart illustrating a grouping process according to an embodiment of the present invention.

[0016] Figure 4 A flowchart illustrating the associated process according to an embodiment of the present invention.

[0017] Figure 5 This is a flowchart illustrating an obstacle detection process according to an embodiment of the present invention.

[0018] Figure 6 This is a schematic diagram illustrating the transmission detection signal according to an embodiment of the present invention.

[0019] Figure 7 This is another flowchart illustrating the grouping process and the association process according to an embodiment of the present invention.

[0020] Figure 8 This is a schematic diagram illustrating the grouping process of this invention for grouping multiple sub-targets.

[0021] Figures 9A to 9CThe figures represent the distance, average received power, and slope of distance versus power estimated by linear regression for a target that is close to a vehicle equipped with a transmitter but is within the transmitter's main beam and not on the impact path, at different time frames.

[0022] Figure 9D This is a schematic diagram illustrating the estimation of the correlation between distance and power using the Pearson correlation coefficient.

[0023] Figure 10 This diagram illustrates the detection of an overhead object approaching a vehicle, as shown in an embodiment of the present invention.

[0024] Figure 11 This is a schematic diagram illustrating the detection of a primary target approaching a vehicle and on the impact path, according to an embodiment of the present invention.

[0025] Figures 12A to 12C The figures represent the distance, average received power, and slope of distance versus power estimated by linear regression for a main target that is close to a vehicle equipped with a transmitter but within the transmitter's main beam and on the impact path, at different time frames.

[0026] Figure 12D This is a schematic diagram illustrating the estimation of the correlation between distance and power using the Pearson correlation coefficient.

[0027] Explanation of key component symbols:

[0028] 1. Object Detection System

[0029] 10 transmitters

[0030] 12 receivers

[0031] 14 Processing Circuit

[0032] 100 First RF front-end circuit

[0033] 120 Second RF front-end circuit

[0034] 122 Analog-to-Digital Converter

[0035] 140 Testing Process

[0036] 142 Grouping Process

[0037] 144 Related Processes

[0038] 146 Obstacle Detection Process

[0039] C0 vehicle

[0040] C1 and C2 clusters

[0041] D1 Main Beam Direction

[0042] D2, D3 direction

[0043] MB Main Beam

[0044] O1, O2 Main Objectives

[0045] Rx receiving antenna

[0046] Tx transmitting antenna Detailed Implementation

[0047] The following specific embodiments illustrate the implementation of the "object detection system and method" disclosed in this invention. Those skilled in the art can understand the advantages and effects of this invention from the content disclosed in this specification. This invention can be implemented or applied through other different specific embodiments, and various details in this specification can also be modified and changed based on different viewpoints and applications without departing from the concept of this invention. Furthermore, the accompanying drawings of this invention are for simple illustrative purposes only and are not depictions of actual dimensions, as stated in advance. The following embodiments will further describe the relevant technical content of this invention in detail, but the disclosed content is not intended to limit the scope of protection of this invention. In addition, the term "or" as used herein should be interpreted to include, depending on the actual situation, any combination of any one or more of the associated listed items.

[0048] Figure 1 This is a functional block diagram illustrating an object detection system according to an embodiment of the present invention. (See also...) Figure 1 As shown, an embodiment of the present invention provides an object detection system 1, which includes a transmitter 10, a receiver 12, and a processing circuit 14.

[0049] The transmitter 10 may include a transmitting antenna Tx and a first radio frequency (RF) front-end circuit 100, and the receiver 12 may include a receiving antenna Rx, a second RF front-end circuit 120, and an analog-to-digital converter 122.

[0050] The first RF front-end circuit 100 and the second RF front-end circuit 120 are used to control the transmitter 10 and the receiver 12, respectively, and can be integrated into one or more chips. In addition, the analog-to-digital converter 122 can be electrically connected between the second RF front-end circuit 120 and the processing circuit 14 to convert analog signals into digital signals for further processing by the processing circuit 14.

[0051] In an embodiment of the present invention, Figure 1The object detection system 1 shown can operate with only a single receiver having a narrow-beam antenna pattern, thereby reducing the need for an RF front-end and an analog-to-digital converter. The narrow beam can be defined by the beamwidth of the antenna pattern; for example, a narrow beam is defined as the angle between two directions on either side of the main beam direction (i.e., the direction of maximum radiation) where the radiated power drops by 3 dB is less than 60 degrees. This narrow-beam characteristic will be used to detect objects in the following embodiments.

[0052] Furthermore, the processing circuit 14 may be, for example, a microcontroller, a microprocessor, or a digital signal processor (DSP), and the processing circuit 14 is connected to the transmitter 10 and the receiver 12.

[0053] In embodiments of the present invention, the processing circuit 14 can be used to execute a detection process 140, a grouping process 142, an association process 144, and an obstacle detection process 146. Generally speaking, the detection process 140 is mainly used to transmit detection signals, receive and process reflected detection signals; the grouping process 142 groups the processed results based on similarity; the association process 144 compares the grouping results and extracts possible primary targets; and the obstacle detection process 146 analyzes whether a warning event is likely to occur with the primary target.

[0054] The following will be referenced together. Figures 2 to 5 To illustrate the object detection method of the present invention, which is applicable to Figure 1 The present invention is an object detection system 1, but is not limited thereto. Figure 2 This is a flowchart illustrating the detection process according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating a grouping process according to an embodiment of the present invention. Figure 4 This is a flowchart illustrating the associated process according to an embodiment of the present invention. Figure 5 This is a flowchart illustrating an obstacle detection process according to an embodiment of the present invention.

[0055] like Figure 2 As shown, the detection process 140 includes configuring the processing circuit 14 to perform the following steps:

[0056] Step S20: Control the transmitter 10 to transmit multiple detection signals in a predetermined pattern toward the main beam direction at different time frames. Here, the different time frames may be, for example, transmitting multiple detection signals cyclically with a predetermined period, or transmitting detection signals based on a variable period.

[0057] Among them, you can refer to Figure 6The figure shows a schematic diagram of a transmission detection signal according to an embodiment of the present invention. As shown, the transmitter 10 can be mounted on the carrier C0, and the main beam direction D1 corresponds to the main beam MB in the predetermined field generated by the transmitter 10 through beamforming. The aforementioned narrow beam characteristic refers to the beamwidth defined by the main beam MB. For example, the angle between the two directions D2 and D3 on both sides of the main beam direction (i.e., the maximum radiation direction) D1, where the radiated power decreases by 3dB, is less than 60 degrees.

[0058] Step S21: Control receiver 12 to receive multiple reflected signals generated by the reflection of these detection signals.

[0059] Step S22: Calculate the corresponding received power, distance, and speed based on the reflected signals. Compared to existing technologies that require multiple antennas and angle detection, in this embodiment of the invention, due to the simplified system architecture, detection process 140 does not measure the incident angle of the reflected signal, but only calculates the received power, distance, and speed corresponding to the reflected signal.

[0060] Furthermore, such as Figure 3 As shown, the grouping process 142 includes configuring the processing circuit 14 to perform the following steps:

[0061] Step S30: Perform a clustering process on the distances and speeds to find the received power, distances and speeds corresponding to at least one primary target.

[0062] In detail, when multiple detection signals are reflected, the multiple reflected signals returned will be received by receiver 12 and processed to generate corresponding information for multiple secondary targets. Therefore, it is necessary to group the information of these secondary targets to find the information of the corresponding primary target.

[0063] Therefore, the clustering process further includes the following steps:

[0064] Step S31: For each time frame, based on the similarity of these velocities and distances, these secondary targets are grouped to find a group of secondary targets corresponding to the primary target.

[0065] Step S32: For each time frame, average the power of these secondary targets to obtain the received power of the primary target.

[0066] Therefore, the aforementioned clustering process 142 will form one or more detection groups based on similar distances and velocities, and each detection group can be regarded as corresponding to a main target. The received power corresponding to this main target will be averaged and stored. Furthermore, since clustering is performed for each time frame, if the main target has different distances in different time frames, it can be determined whether the main target is moving relative to the receiver 12 (or the vehicle equipped with the receiver).

[0067] Furthermore, such as Figure 4 As shown, the associated process 144 includes configuring the processing circuit 14 to perform the following steps:

[0068] Step S40: Execute the association process 144 to track the distance, received power, and speed of the main target in different time frames.

[0069] In detail, the association process 144 tracks a new set of secondary targets in the current time frame and compares them with the set of secondary targets detected in the previous time frame. The association process 144 can pair the set of secondary targets in the current time frame with the most likely set of secondary targets in the previous time frame based on range and velocity. In other words, the association process 144 is primarily used to track the trajectory of the primary target, enabling the radar system to track the primary target based on received range, velocity, angle, and received power.

[0070] Therefore, the associated process 144 may further include the following steps:

[0071] Step S41: Compare the sub-targets grouped under different time frames.

[0072] Step S42: Pair the groups in the second time frame and the first time frame based on speed and distance.

[0073] Step S43: Track the secondary targets corresponding to the primary target in the first time frame and the second time frame.

[0074] Based on the above explanation, judgment criteria can be added to the grouping process 142 and the association process 144, which can be further referenced. Figure 7 , Figure 7 This is another flowchart illustrating the grouping process 142 and the association process 144 according to an embodiment of the present invention.

[0075] like Figure 7 As shown, the grouping process 142 and the association process 144 may include the following steps:

[0076] Step S700: Initialize the distance threshold and velocity threshold, and begin setting the clustering parameters for all sub-targets (e.g., from 1 to N).

[0077] Step S701: Set the count value i = 1.

[0078] Step S702: Set the count value j = i + 1.

[0079] Step S703: Determine whether the absolute difference between the distance of the i-th sub-target and the distance of the j-th sub-target is less than the distance threshold.

[0080] In response to the absolute difference between the distance of the i-th sub-target and the distance of the j-th sub-target being less than the distance threshold, proceed to step S704: determine whether the absolute difference between the velocity of the i-th sub-target and the velocity of the j-th sub-target is less than the velocity threshold.

[0081] If the absolute difference between the velocity of the i-th sub-target and the velocity of the j-th sub-target is less than the velocity threshold, proceed to step S705: group the i-th sub-target and the j-th sub-target into the same group.

[0082] Step S706: Set the count value j = j + 1.

[0083] In response to the absolute difference between the distance of the i-th sub-target and the distance of the j-th sub-target being not less than a distance threshold in step S703, and in response to the absolute difference between the velocity of the i-th sub-target and the velocity of the j-th sub-target being not less than a velocity threshold in step S704, proceed to step S706.

[0084] Step S707: Determine whether j is greater than N.

[0085] If j is not greater than N, return to step S702. If j is greater than N, proceed to step S708: set the count value i = i + 1.

[0086] Step S709: Determine if i is greater than N.

[0087] If i is not greater than N, return to step S701.

[0088] When i is greater than N, the clustering process ends, and the process proceeds to step S710. It should be noted that the groups following the clustering process correspond to the main target mentioned in the previous steps.

[0089] In detail, in the above steps, due to the limited range resolution of the radar, multiple reflected signals may be received for the same object. In other words, these reflected signals correspond to multiple secondary targets, but may originate from the same primary target. In other words, by setting a range threshold, if the distance between different secondary targets (even if they come from different objects) is less than the range threshold, they are grouped together, that is, the two secondary targets are considered to originate from the same primary target.

[0090] On the other hand, even in the presence of multiple targets, multiple main targets can be identified through the grouping process 142 and the association process 144, and then it can be determined in the subsequent obstacle detection process 146 whether a warning event will occur with any detected main target.

[0091] The above process explains how different detected secondary targets form a cluster based on distance and velocity. If different secondary targets have matching distances and velocities, they can be considered to come from the same primary target.

[0092] Further reference is available. Figure 8 This is a schematic diagram illustrating the grouping process of this invention to group multiple sub-targets. For example... Figure 8 As shown in the figure, two objects and multiple detected secondary targets are displayed. "+" indicates a secondary target detected from the first object, and "*" indicates a secondary target detected from the second object. The difference between the two objects lies in their velocities. From time 1 to time 2, it can be seen that the first object moves farther than the second object.

[0093] By executing steps S700 to S709, the distance and speed of all detected secondary targets can be compared, and then different clusters can be formed.

[0094] For example, targets of the same class "+" will be grouped into the same cluster C1 because they have similar distances and speeds, while targets of the same class "*" will be grouped into the same cluster C2 because they have similar distances and speeds. In this embodiment, only distance and speed information is needed to complete the above grouping process; angle information is not required.

[0095] However, in order to further track the main target that has been detected, that is, to track the clustering results through the correlation process, the following steps need to be performed.

[0096] Step S710: For all sub-targets, set the associated power value Power of the same cluster to the maximum power value in that cluster.

[0097] For example, suppose that three sub-targets are detected from an object, with distances r1, r2, r3, powers p1, p2, p3, and velocities v1, v2, v3, respectively. The distances are in the order r1 > r2 > r3, and the powers are in the order p2 > p1 > p3.

[0098] The purpose of step S710 is to represent the cluster with the maximum power among the three sub-targets. In other words, the associated power value ASS_Power of this cluster is set to the maximum power p2, because the maximum power usually represents that the sub-target has a relatively high accuracy.

[0099] Step S711: For all sub-targets, set the associated distance value ASS_Range of the same cluster to the minimum distance in that cluster. For example, set the associated distance value ASS_Range of this cluster to the shortest distance r1. The shortest distance ensures that the detected object is the closest distance to the detection point when determining whether a warning event will occur.

[0100] Step S712: For all sub-targets, set the associated velocity value ASS_Velocity of the same cluster to the average velocity of all sub-targets in that cluster. For example, set the associated velocity value ASS_Velocity of this cluster to velocity (v1+v2+v3) / 3.

[0101] Step S713: Based on the association distance value and association velocity value of the cluster obtained in the previous time frame, estimate the predicted distance of the current time frame using the association formula for tracking the cluster.

[0102] That is, the relationship can be expressed by the following formula (1):

[0103] Range Predicted =Range Previous +ASS_Velocity*Time cycle …Formula (1);

[0104] Among them, Range Predicted Range is the predicted distance for the current time frame. Previous The distance from the previous time frame, Time cycle ASS_Velocity is the length of the time frame, and ASS_Velocity is the average velocity value of the previous time frame.

[0105] Step S714: If the predicted distance of the current time frame and the associated velocity value of the previous time frame match the associated distance and associated velocity values ​​of any cluster obtained in the current time frame, then the processing circuit updates the information of the matching cluster to the associated distance value, associated power value, and associated velocity value. The update involves the processing circuit 14 updating the information of clusters that were already determined to match or associated in the previous time frame (e.g., the associated distance value, associated power value, and associated velocity value of the previous time frame) to the associated distance value, associated power value, and associated velocity value of the clusters determined to match in the current time frame.

[0106] For more details, please refer to Figure 8At time 1, the power, distance, and speed of the cluster were calculated, and the above steps S710, S711, and S712 were performed to form cluster C1 formed by the sub-target "+" and cluster C2 formed by the sub-target "*" as defined at time 1.

[0107] Furthermore, after executing step S713, the predicted distance and speed for estimating clusters C1 and C2 can be calculated.

[0108] At time 2, similarly, the power, distance, and speed of the clusters are calculated, and through the above steps S710, S711, and S712, cluster C1 formed by the sub-target "+" and cluster C2 formed by the sub-target "*" defined at time 2 are formed.

[0109] Next, the predicted distance and velocity estimated at time 1 can be compared with the distance (associated distance value Ass_Range) and velocity (associated velocity value Ass_Velocity) of each cluster detected at time 2. If they match, new distance, velocity, and power are obtained for the corresponding cluster.

[0110] Please refer to this again. Figure 5 After the association process 144, the object detection method of the present invention enters the obstacle detection process 146. For example... Figure 5 As shown, the obstacle detection process 146 includes a configuration processing circuit 14 to perform the following steps:

[0111] Step S50: Calculate the power change trend and distance change trend of the main target.

[0112] Step S51: Determine whether an early warning event will occur based on the relationship between the power change trend and the distance change trend.

[0113] In response to the distance change trend indicating that the main target is approaching, and the power change trend indicating that the received power of the corresponding main target is increasing, proceed to step S52: determine that a warning event will occur with the main target.

[0114] If the distance change trend indicates that the main target is approaching, and the power change trend indicates that the received power of the corresponding main target is decreasing, then proceed to step S53: determine that no warning event will occur.

[0115] Furthermore, in response to the distance change trend indicating that the main target is moving away, proceed to step S53: determine that no warning event will occur.

[0116] Specifically, in step S50, the processing circuit 14 can be configured to calculate the relationship between the power change trend and the distance change trend using linear regression, logistic regression, lasso regression, or classification algorithms.

[0117] Taking linear regression as an example, linear regression is a technique for modeling the relationship between dependent variables. This modeling technique can be used to determine the relationship between the received power of the main target and the distance.

[0118] The linear equation can be expressed as y(power) = ax(distance) + b... (Equation 1);

[0119] For y = ax + b, if the given observation values ​​are {(x1, y1), (x2, y2), ..., (xN, yN)}, when linear regression is used in this embodiment, xn and yn are the distance to the main target detected and tracked within the nth time frame of the radar's field of view, and the average received power.

[0120] The error for each observation and its linear prediction is:

[0121] {(y1-(ax1+b)), (y2-(ax2+b)),…,(y N -(ax N +b))}.

[0122] The mean square of the prediction error E can be written as:

[0123]

[0124]

[0125] Minimizing E(a, b) relative to a will produce the following equation (4):

[0126]

[0127]

[0128]

[0129]

[0130] Minimizing E(a, b) with respect to b will produce the following equation:

[0131]

[0132]

[0133]

[0134]

[0135] By solving equations (7) and (11), the values ​​of a and b can be found:

[0136]

[0137]

[0138] Equations (12) and (13) can also be written as:

[0139]

[0140]

[0141] Equations (12) and (13) show how the slope and intercept of distance and power measurements are collected in one go from N time frames. This operation can be performed frame-by-frame using equations (14) and (15), for example, acquiring new distance and power information for each time frame and calculating the slope and intercept.

[0142] The slope is calculated in each time frame, not at the end. If the slope 'a' is positive, the primary target will not collide with the vehicle equipped with launcher 10, and no alarm will be triggered. If the slope 'a' is negative, an alarm will be triggered.

[0143] Please refer to further information. Figures 9A to 9C as well as Figure 10 , Figures 9A to 9C The figures represent the distance, average received power, and slope variation curves of distance versus power estimated by linear regression for a main target that is close to a vehicle equipped with transmitter 10 but is within the main beam of transmitter 10 and not on the impact path, at different time frames. Figure 10 The illustration, based on an embodiment of the present invention, shows a detection schematic of a main target O1 being an overhead object approaching the vehicle C0.

[0144] exist Figure 10 In such cases, Figure 9A As shown, the distance to the main target decreases over time; however, as... Figure 9B As shown, the average received power of the main target also decreases over time. Further estimation using linear regression shows that the slope is positive. Therefore, this slope indicates that the distance and power have the same trend. However, the power change trend indicates that the received power of the corresponding main target is decreasing, which means that the main target is moving away from the main beam of the transmitter 10. Therefore, no warning event will occur.

[0145] In addition to using linear regression to estimate the relationship between power change trends and distance change trends, the Pearson correlation coefficient can also be used to estimate this relationship.

[0146] In an embodiment of the present invention, the Pearson correlation coefficient r equation can be rewritten as shown in equation (16):

[0147]

[0148] Equation (16) above is used to determine the relationship between variables. The aforementioned linear regression determines the relationship between variables by obtaining the slope; however, the correlation r... xy It can also explain how explanatory variables are relative, and whether the variables needed to determine the main objective are directly related or inversely related.

[0149] The correlation coefficient ranges from -1 to 1. A correlation coefficient of 1 indicates that the linear equation perfectly describes the relationship between the two variables X and Y, with all data points lying on a straight line where Y increases as X increases. A correlation coefficient of -1 indicates that all data points lie on a separate straight line where Y decreases as X increases.

[0150] When the correlation coefficient is 0, it means that there is no linear relationship between the variables.

[0151] The following will be aimed at Figure 10 In one embodiment, Pearson's correlation coefficient is used to determine the relationship between the trends of distance change and power change.

[0152] Please refer to Figure 9D This is a schematic diagram illustrating the estimation of the correlation between distance and power using the Pearson correlation coefficient. Wherein, Figure 9D The estimated Pearson correlation coefficient r xy (Where the horizontal axis x represents distance and the vertical axis y represents power) is 0.8792. In other words, objects not in the collision path will have a positive correlation value, and this positive correlation indicates that distance and power have the same trend of change. This conclusion is consistent with... Figure 9C They are the same.

[0153] Please refer to further information. Figure 11 , Figures 12A to 12C . Figure 11 This is a schematic diagram illustrating the detection of the main target O2 approaching the vehicle and along the impact path, according to an embodiment of the present invention. Figures 12A to 12C They are respectively Figure 11As shown, the distance, average received power, and slope variation curves of distance versus power estimated by linear regression are shown for the main target O2, which is close to the vehicle C0 equipped with transmitter 10 but within the main beam MB of transmitter 10 and on the impact path, at different time frames.

[0154] exist Figure 11 In such cases, Figure 12A As shown, the distance to the main target decreases over time, and as... Figure 12B As shown, the average received power of the main target increases over time. Further estimation using linear regression reveals a negative slope. Therefore, this slope indicates that the distance and power have different trends. Since the distance decreases and the power increases, it means that the main target is approaching the vehicle C0 along the main beam direction of the transmitter 10. Therefore, it is determined that a warning event will occur.

[0155] Similarly, please refer to Figure 12D This is a schematic diagram illustrating the estimation of the correlation between distance and power using the Pearson correlation coefficient. Wherein, Figure 12D The estimated Pearson correlation coefficient r xy (Where the horizontal axis x represents distance and the vertical axis y represents power) is -0.83194. In other words, objects in the collision path will have a negative correlation value, and this negative correlation indicates that distance and power have different trends of change. This conclusion is consistent with... Figure 12C They are the same.

[0156] [Beneficial Effects of the Examples]

[0157] One of the advantages of the present invention is that the object detection system and object detection method provided by the present invention will allow the trajectory of the object to be inferred without additional hardware modifications, and are applicable to situations where it is not necessary to know the angle of the object relative to the transmitter with very high precision, as well as situations involving multiple target objects.

[0158] This invention does not require highly precise determination of the object's angle. Instead, it determines the relationship between distance and power change trends by calculating information such as the distance of the reflected signal, the distance between the object and the receiver, and the received power. When it is only necessary to guess whether an object constitutes an obstacle, the required number of antennas is smaller, resulting in fewer corresponding RF front-end circuits and analog-to-digital converters. Therefore, the object detection system and method of this invention are superior to existing systems in terms of cost and complexity.

[0159] The above-disclosed content is only a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of the claims of the present invention. Therefore, all equivalent technical changes made based on the description and drawings of the present invention are included within the scope of the claims of the present invention.

Claims

1. An object detection system, the detection system comprising: One transmitter; One receiver; as well as A processing circuit, connected to the transmitter and the receiver, is configured to: The transmitter is controlled to transmit multiple detection signals in different time frames in a predetermined field pattern toward a main beam direction, wherein the main beam direction corresponds to a main beam in the predetermined field pattern generated by the transmitter through beamforming. The receiver is controlled to receive multiple reflected signals generated by the reflection of these detection signals; Based on these reflected signals, calculate the corresponding received power, distance, and speed. Perform a clustering process on these distances and these speeds to find the received power, these distances and these speeds corresponding to at least one primary target; Perform an association process to track the distances and received powers of the at least one primary target in different time frames; as well as Calculate a power change trend and a distance change trend for the at least one primary target, and determine whether an early warning event will occur based on the relationship between the power change trend and the distance change trend. In response to the distance change trend indicating that the at least one main target is approaching, and the power change trend indicating that the received power corresponding to the at least one main target is increasing, it is determined that the warning event will occur with the at least one main target. These reflected signals correspond to multiple sub-targets, and the grouping process includes: For each time frame, based on the similarity of the velocities and distances, the sub-targets are grouped to find at least one cluster of the sub-targets corresponding to the at least one primary target; For each time frame, these secondary targets are averaged to obtain the received power of the at least one primary target; The associated process includes: Compare these sub-targets that have been grouped at different time frames; Based on these speeds and these distances, the at least one cluster segmented in a second time frame and a first time frame is paired; and Track the at least one primary target and the corresponding secondary targets in the first time frame and the second time frame, respectively; For these sub-targets, obtain a first associated distance value and a first associated velocity value for at least one cluster in the first time frame; Based on the first associated distance value and the first associated velocity value of the at least one cluster obtained in the first time frame, a predicted distance of the at least one cluster in the second time frame is estimated using an association formula. For these sub-targets, obtain a second associated distance value and a second associated velocity value for at least one cluster in the second time frame; and The processing circuit determines whether the predicted distance and the first associated speed value match the second associated distance value and the second associated speed value. If so, the processing circuit updates the information of the at least one matching cluster with the second associated distance value and the second associated speed value.

2. The detection system of claim 1, wherein the transmitter includes a first radio frequency front-end circuit, and the receiver includes a second radio frequency front-end circuit and an analog-to-digital converter.

3. The detection system of claim 1, wherein the processing circuit is further configured to calculate the relationship between the power change trend and the distance change trend by means of linear regression, logistic regression, lasso regression or classification algorithm.

4. The detection system of claim 1, wherein the power change trend of the at least one primary target is calculated by averaging multiple average powers based on the received power corresponding to the tracked secondary targets in different time frames.

5. The detection system as described in claim 1, wherein the associated process includes: For these sub-targets, obtain a first associated power value for at least one cluster in the first time frame; For these sub-targets, obtain a second associated power value for at least one cluster in the second time frame; and When the predicted distance and the first associated speed value are determined to match the second associated distance value and the second associated speed value, the processing circuit updates the information of the at least one matching cluster to the second associated power value.

6. The detection system of claim 5, wherein the step of obtaining the first associated power value, the first associated distance value, and the first associated velocity value of the at least one cluster includes: The first associated power value of the at least one cluster is set to the maximum power value of the sub-targets in the at least one cluster in the first time frame; For these sub-targets, the first associated distance value of the at least one cluster is set to the minimum distance among the sub-targets in the at least one cluster in the first time frame; as well as For these sub-targets, the first associated velocity value of the at least one cluster is set to the average velocity of the sub-targets in the at least one cluster in the first time frame.

7. The detection system as described in claim 1, wherein, In response to the distance change trend indicating that at least one primary target is approaching, and the power change trend indicating that the received power corresponding to the at least one primary target is decreasing, it is determined that the warning event will not occur. as well as If the distance change trend indicates that at least one primary target is moving away, it is determined that the warning event will not occur.

8. An object detection method applicable to a detection system, the detection system including a transmitter, a receiver, and a processing circuit, the detection method including configuring the processing circuit to: The transmitter is controlled to transmit multiple detection signals in different time frames in a predetermined field pattern toward a main beam direction, wherein the main beam direction corresponds to a main beam in the predetermined field pattern generated by the transmitter through beamforming. The receiver is controlled to receive multiple reflected signals generated by the reflection of these detection signals; Based on these reflected signals, calculate the corresponding received power, distance, and speed. Perform a clustering process on these distances and these speeds to find the received power, these distances and these speeds corresponding to at least one primary target; Perform an association process to track the distances and received powers of the at least one primary target in different time frames; as well as Calculate a power change trend and a distance change trend for the at least one primary target, and determine whether an early warning event will occur based on the relationship between the power change trend and the distance change trend. in, In response to the distance change trend indicating that the at least one primary target is approaching, and the power change trend indicating that the received power corresponding to the at least one primary target is increasing, it is determined that the warning event will occur with the at least one primary target; These reflected signals correspond to multiple sub-targets, and the grouping process includes: For each time frame, based on the similarity of the velocities and distances, the sub-targets are grouped to find at least one cluster of the sub-targets corresponding to the at least one primary target; For each time frame, these secondary targets are averaged to obtain the received power of the at least one primary target; The associated process includes: Compare these sub-targets that have been grouped at different time frames; Based on these speeds and these distances, the at least one cluster segmented in a second time frame and a first time frame is paired; and Track the at least one primary target and the corresponding secondary targets in the first time frame and the second time frame, respectively; For these sub-targets, obtain a first associated distance value and a first associated velocity value for at least one cluster in the first time frame; Based on the first associated distance value and the first associated velocity value of the at least one cluster obtained in the first time frame, a predicted distance of the at least one cluster in the second time frame is estimated using an association formula. For these sub-targets, obtain a second associated distance value and a second associated velocity value for at least one cluster in the second time frame; and Determine whether the predicted distance and the first associated velocity value match the second associated distance value and the second associated velocity value. If so, update the information of the at least one matching cluster with the second associated distance value and the second associated velocity value.

9. The detection method of claim 8, wherein the transmitter includes a first radio frequency front-end circuit, and the receiver includes a second radio frequency front-end circuit and an analog-to-digital converter.

10. The detection method of claim 8, wherein the processing circuit is further configured to calculate the relationship between the power change trend and the distance change trend by means of linear regression, logistic regression, lasso regression or classification algorithm.

11. The detection method of claim 8, wherein the power change trend of the at least one primary target is calculated by averaging multiple average powers based on the received power corresponding to the tracked secondary targets in different time frames.

12. The detection method as described in claim 8, wherein the correlation process includes: For these sub-targets, obtain a first associated power value for at least one cluster in the first time frame; For these sub-targets, obtain a second associated power value for at least one cluster in the second time frame; and When it is determined that the predicted distance and the first associated velocity value match the second associated distance value and the second associated velocity value, the information of the at least one matching cluster is updated to the second associated power value.

13. The detection method of claim 12, wherein the step of obtaining the first associated power value, the first associated distance value, and the first associated velocity value of the at least one cluster includes: The first associated power value of the at least one cluster is set to the maximum power value of the sub-targets in the at least one cluster in the first time frame; For these sub-targets, the first associated distance value of the at least one cluster is set to the minimum distance among the sub-targets in the at least one cluster in the first time frame; as well as For these sub-targets, the first associated velocity value of the at least one cluster is set to the average velocity of the sub-targets in the at least one cluster in the first time frame.

14. The detection method as described in claim 8, wherein, In response to the distance change trend indicating that at least one primary target is approaching, and the power change trend indicating that the received power corresponding to the at least one primary target is decreasing, it is determined that the warning event will not occur. as well as If the distance change trend indicates that at least one primary target is moving away, it is determined that the warning event will not occur.

Images