Target object detection method, device, computer equipment and storage medium

Abstract

The present application relates to a target object detection method, apparatus, computer device, storage medium, and computer program product. The method comprises: obtaining an echo mean data set, the echo mean data set comprising multiple echo mean values, the echo mean values being determined by echo signals reflected from the interior of a target vehicle to be detected; performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity representing the number of envelope features contained in the echo mean data set, and the second feature quantity representing the number of burr features contained in each envelope feature in the echo mean data set; and determining a target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and identification strategy. This method improves the accuracy of target object detection.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a target object detection method, apparatus, computer equipment, storage medium, and computer program product. Background Art

[0002] As people's awareness of automobile safety performance continues to improve, in-car safety has also received more and more attention. Therefore, in order to avoid the danger of accidentally leaving children, pets and other target objects in the car, in-car object detection technology has emerged.

[0003] Current methods for detecting target objects use radar installed inside a vehicle to detect the presence of a target object. Specifically, if the radar detects a change in the reflected electromagnetic wave signal, the target object is determined to be present. Otherwise, the target object is considered absent.

[0004] However, current target object detection methods rely on changes in electromagnetic wave signals reflected by objects inside the vehicle to determine whether there is a target object inside the vehicle. These methods are easily interfered with by the swinging of non-target objects, resulting in false positives or missed positives in the detection of target objects inside the vehicle, and inaccurate detection results. Summary of the Invention

[0005] Based on this, it is necessary to provide a target object detection method, device, computer equipment, computer-readable storage medium and computer program product to address the above technical problems.

[0006] In a first aspect, the present application provides a method for detecting a target object. The method comprises:

[0007] Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected;

[0008] performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features contained in the echo mean data set, and the second feature quantity represents the number of burr features contained in each envelope feature of the echo mean data set;

[0009] A target object detection result corresponding to the target vehicle to be detected is determined based on the first feature quantity, the second feature quantity and a preset detection and identification strategy.

[0010] In one embodiment, obtaining the echo mean value dataset includes:

[0011] receiving an echo signal reflected from the interior of a target vehicle to be detected, wherein each frame of the echo signal includes an echo data sequence;

[0012] performing mean processing on the echo data in each frame of the echo data sequence to obtain an echo mean;

[0013] Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each of the echo mean values to obtain an echo mean value data set.

[0014] In one embodiment, extracting features from the echo mean value dataset according to a preset feature extraction algorithm to determine the first feature quantity and the second feature quantity includes:

[0015] extracting envelope features contained in the echo mean data set according to a preset first feature extraction algorithm and the echo mean data set, and performing statistical processing on the envelope features to determine a first feature quantity;

[0016] The burr feature contained in the envelope feature is extracted according to a preset second feature extraction algorithm, and the burr feature is statistically processed to determine a second feature quantity.

[0017] In one embodiment, extracting the envelope features contained in the echo mean value dataset according to a preset first feature extraction algorithm, and performing statistical processing on the envelope features to determine the first feature quantity includes:

[0018] traversing each of the echo mean values in the echo mean value data set, and identifying peak data in the echo mean value data set;

[0019] For each peak data, obtaining a second preset number of echo means including the peak data and frame number identifiers corresponding to the echo means, and determining whether the echo means and the frame number identifiers corresponding to the echo means are an envelope feature according to a preset envelope feature discrimination condition;

[0020] Quantitative statistics are performed on the envelope features contained in the echo mean value data set to obtain a first feature quantity.

[0021] In one embodiment, determining the target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and recognition strategy includes:

[0022] When the first feature quantity is greater than a preset first feature quantity threshold, determining a target object detection result corresponding to the target vehicle to be detected according to the second feature quantity and the preset second feature quantity threshold;

[0023] When the first feature quantity is less than or equal to the first feature quantity threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

[0024] In one embodiment, the second feature value threshold includes a first sub-threshold and a second sub-threshold of the second feature value, the first sub-threshold is greater than the second sub-threshold, and determining the target object detection result corresponding to the target vehicle to be detected based on the second feature value and the preset second feature value threshold includes:

[0025] When all corresponding second feature quantities in the echo mean value data set are less than or equal to the first sub-threshold, if all the second feature quantities are greater than the second sub-threshold, determining that the target object detection result corresponding to the target vehicle to be detected is that there is no target object inside the vehicle;

[0026] When a second feature value exists in the echo mean value data set and is greater than the first sub-threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

[0027] In a second aspect, the present application further provides a target object detection device. The device comprises:

[0028] an acquisition module, configured to acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected;

[0029] a first determining module, configured to perform feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features included in the echo mean data set, and the second feature quantity represents the number of burr features included in each envelope feature of the echo mean data set;

[0030] The second determination module is used to determine the target object detection result corresponding to the target vehicle to be detected based on the first feature value, the second feature value and a preset detection and recognition strategy.

[0031] In one embodiment, the acquisition module is used to receive an echo signal reflected from the interior of the target vehicle to be detected, and each frame of the echo signal contains an echo data sequence;

[0032] performing mean processing on the echo data in each frame of the echo data sequence to obtain an echo mean;

[0033] Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each of the echo mean values to obtain an echo mean value data set.

[0034] In one embodiment, the first determining module is configured to extract envelope features contained in the echo mean value dataset according to a preset first feature extraction algorithm, and perform statistical processing on the envelope features to determine a first feature quantity;

[0035] The burr feature contained in the envelope feature is extracted according to a preset second feature extraction algorithm, and the burr feature is statistically processed to determine a second feature quantity.

[0036] In one embodiment, the first determining module is configured to traverse each echo mean in the echo mean data set and identify peak data in the echo mean data set;

[0037] For each peak data, obtaining a second preset number of echo means including the peak data and frame number identifiers corresponding to the echo means, and determining whether the echo means and the frame number identifiers corresponding to the echo means are an envelope feature according to a preset envelope feature discrimination condition;

[0038] Quantitative statistics are performed on the envelope features contained in the echo mean value data set to obtain a first feature quantity.

[0039] In one embodiment, the second determining module is configured to determine a target object detection result corresponding to the target vehicle to be detected based on the second feature quantity and the preset second feature quantity threshold when the first feature quantity is greater than a preset first feature quantity threshold;

[0040] When the first feature quantity is less than or equal to the first feature quantity threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

[0041] In one embodiment, the second feature value threshold includes a first sub-threshold and a second sub-threshold of the second feature value, the first sub-threshold is greater than the second sub-threshold, and the second determination module is configured to, when all corresponding second feature values in the echo mean value dataset are less than or equal to the first sub-threshold, determine that the target object detection result corresponding to the target vehicle to be detected is that there is no target object inside the vehicle if all the second feature values are greater than the second sub-threshold;

[0042] When a second feature value exists in the echo mean value data set and is greater than the first sub-threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

[0043] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are performed:

[0044] Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected;

[0045] performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features contained in the echo mean data set, and the second feature quantity represents the number of burr features contained in each envelope feature of the echo mean data set;

[0046] A target object detection result corresponding to the target vehicle to be detected is determined based on the first feature quantity, the second feature quantity and a preset detection and identification strategy.

[0047] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0048] Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected;

[0049] performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features contained in the echo mean data set, and the second feature quantity represents the number of burr features contained in each envelope feature of the echo mean data set;

[0050] A target object detection result corresponding to the target vehicle to be detected is determined based on the first feature quantity, the second feature quantity and a preset detection and identification strategy.

[0051] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the following steps:

[0052] Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected;

[0053] performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features contained in the echo mean data set, and the second feature quantity represents the number of burr features contained in each envelope feature of the echo mean data set;

[0054] A target object detection result corresponding to the target vehicle to be detected is determined based on the first feature quantity, the second feature quantity and a preset detection and identification strategy.

[0055] In the target object detection method, apparatus, computer device, storage medium, and computer program product, a radar acquires an echo mean dataset. The echo mean dataset contains multiple echo mean values. The echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected. The radar extracts features from the echo mean dataset using a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity. Based on the first feature quantity, the second feature quantity, and a preset detection and identification strategy, the radar determines a target object detection result corresponding to the target vehicle to be detected. This method uses a preset feature extraction algorithm to determine the number of envelope features and burr features (i.e., the first feature quantity and the second feature quantity) in the echo mean dataset. The number of envelope features and burr features is analyzed and processed according to a preset detection and identification strategy. This method effectively identifies target objects (i.e., living objects) and non-target objects (i.e., non-living objects), thereby avoiding false positives caused by the movement of non-living objects within the vehicle and missed positives caused by living objects. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 1 is a flow chart of a target object detection method according to an embodiment;

[0057] Figure 2 Schematic diagram of a flow chart of the steps of obtaining an echo mean value dataset in one embodiment;

[0058] Figure 3 A schematic flow chart of the steps of determining the first characteristic value and the second characteristic value in one embodiment;

[0059] Figure 4 FIG. 1 is a flow chart of a step of determining a first feature value in one embodiment;

[0060] Figure 5 A schematic diagram of a process for determining a target object detection result in one embodiment;

[0061] Figure 6 A schematic diagram of a flow chart of steps for determining a target object detection result in another embodiment;

[0062] Figure 7A schematic diagram of a process of detecting a target object according to an embodiment;

[0063] Figure 8 is a structural block diagram of a target object detection device in one embodiment;

[0064] Figure 9 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0066] In one embodiment, Figure 1 As shown, a target object detection method is provided, and the method is described by taking the application of a radar as an example. The radar may include a processing unit (e.g., a CPU, Central Processing Unit) and a data storage unit for data processing. Specifically, the target object detection method may include the following steps:

[0067] Step 102: Acquire an echo mean value dataset.

[0068] The echo mean value data set includes multiple echo mean values, and the echo mean value is determined by the echo signal reflected from the interior of the target vehicle to be detected.

[0069] In practice, a radar is installed inside the target vehicle to be detected. The radar transmits electromagnetic wave signals to detect the target object inside the target vehicle to be detected. Then, after being reflected by various objects inside the target vehicle to be detected, the radar can receive the reflected electromagnetic wave signals (also known as echo signals). Then, ADC sampling (also known as Nyquist sampling) is performed on the echo signals. Then, for the collected echo signals, the radar calculates the echo mean based on the echo data sequence contained in each frame of the echo signal. In addition, the radar is provided with a first preset number (for example, the first preset number is represented by M, M is any non-zero natural number and M is greater than 3) of data acquisition frames (also known as a fixed sampling window). Based on this data acquisition frame, the radar obtains an echo mean data set. Each echo mean data set contains M echo means.

[0070] Optionally, the radar's integrated echo mean data processing unit can also be independent of the radar. For example, the echo mean data processing unit can be integrated into a vehicle-mounted terminal. After the radar collects the echo signal, it sends it to the vehicle-mounted terminal. The processing unit in the vehicle-mounted terminal processes the echo data in the echo signal to obtain a target object detection result. Therefore, the embodiments of the present application do not limit the execution device of the target object detection method.

[0071] Optionally, the radar installed in the target vehicle to be detected can be a millimeter wave radar, which transmits millimeter waves to detect the target object. Due to the strong ability of millimeter waves to penetrate fog, smoke, and dust, it can achieve multi-dimensional resolution of distance, velocity azimuth, and pitch angle, and the millimeter wave radar has a wide field of view (FOV), a large effective bandwidth, and high detection accuracy. Therefore, the radars mentioned below can all be replaced by millimeter wave radars to more accurately detect the target object, and this application does not contain any details.

[0072] Optionally, the radar installed inside the target vehicle to be detected can have a dormant period and an operating period. During normal vehicle operation, the radar can enter the dormant period to reduce resource consumption. When the vehicle stops or is locked, the radar is powered on and activated based on preset activation conditions (or preconditions), causing the radar to enter the operating period from the dormant period and perform real-time detection of the target object inside the vehicle during the operating period. The preset activation conditions for the radar can be, but are not limited to, the vehicle stopping or the vehicle being locked, and are not limited in this embodiment of the application.

[0073] Step 104 : extracting features from the echo mean value dataset according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity.

[0074] In implementation, for the echo mean data set (containing M echo means) collected by a fixed sampling window, the radar can extract the envelope features and burr features contained in the echo mean data set according to a preset feature extraction algorithm. Then, the radar performs statistical processing on the extracted envelope features and burr features to obtain a first feature quantity representing the number of envelope features and a second feature quantity representing the number of burr features, respectively.

[0075] Among them, the envelope feature characterizes the data fluctuation trend of the echo mean data in the echo mean data set. Furthermore, the first feature quantity can be obtained by counting the envelope feature, and the first feature quantity represents the number of envelope features contained in an echo mean data set. The burr feature (represented by burr) characterizes the slight deviation fluctuation of the echo mean data contained in each envelope feature. Furthermore, the second feature quantity can be obtained by counting the burr feature, and the second feature quantity represents the number of burr features contained in an envelope feature. By analyzing the first feature quantity and the second feature quantity, information such as micro-movement or large-scale movement inside the target vehicle to be detected can be obtained. Specifically, when the first feature quantity and the second feature quantity meet the preset quantity conditions, it represents that there is a target object inside the target vehicle to be detected that produces high-frequency micro-movement or low-frequency large-scale movement.

[0076] Step 106 : Determine a target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and recognition strategy.

[0077] In practice, the radar is provided with feature value thresholds for the first feature value and the second feature value, as well as a target object detection and identification strategy within the vehicle. Thus, the radar can determine a target object detection result corresponding to the target vehicle to be detected based on the first feature value, the second feature value, and the preset detection and identification strategy. The target object detection result can include the presence of a target object within the target vehicle to be detected and the absence of a target object within the target vehicle to be detected. Each target object detection result can be further subdivided, i.e., the presence of a target object can be divided into a target object with slight movements within the vehicle and a target object with large movements within the vehicle, etc., which is not limited in the present embodiment of the application.

[0078] Optionally, the radar generates a corresponding control signal for different target object detection results, and sends the control signal to the vehicle-mounted terminal. The vehicle-mounted terminal pre-stores a variety of vehicle control strategies, and then executes the corresponding target vehicle control strategy based on the control signal. For example, the target object in this application can be, but is not limited to, living objects such as children, the elderly, and pets. When the target object detection result is that the target object exists inside the target vehicle to be detected, the radar can send a control signal generated based on the target object detection result to the vehicle-mounted terminal. According to the vehicle early warning and control strategy preset in the vehicle-mounted terminal, the vehicle-mounted terminal generates a control instruction in response to the control signal and sends it to the alarm device, and the alarm device generates an alarm prompt information.

[0079] Optionally, there can be multiple output forms of the alarm prompt information. The alarm prompt information can be output and displayed through the display interface of the vehicle terminal, or through the vehicle speaker, etc. The output forms of multiple alarm prompt information can also be combined to output and display the alarm prompt information. This embodiment does not limit this.

[0080] Optionally, the feature value thresholds of the above-mentioned feature value parameters (ie, the first feature value and the second feature value) are determined through a large number of experiments and can be set based on specific business needs, which is not limited in the embodiments of the present application.

[0081] In the above-mentioned target object detection method, the radar obtains an echo mean data set. The echo mean data set includes multiple echo means. The echo mean is determined by the echo signal reflected from the interior of the target vehicle to be detected. The radar determines a first feature quantity and a second feature quantity based on a preset feature extraction algorithm and the echo mean data set. Based on the first feature quantity, the second feature quantity, and a preset detection and identification strategy, the radar determines a target object detection result corresponding to the target vehicle to be detected. This method, by processing the echo mean data set determined from the echo signal using a preset feature extraction algorithm and detection and identification strategy, can distinguish between target objects and non-target objects, thereby improving the accuracy of target object detection.

[0082] In one embodiment, Figure 2 As shown, the specific processing process of step 102 includes:

[0083] Step 202: Receive the echo signal reflected from the interior of the target vehicle to be detected, wherein each frame of the echo signal contains an echo data sequence.

[0084] In practice, a radar (for example, a millimeter-wave radar) includes a transmitting antenna and a receiving antenna. The transmitting antenna continuously transmits electromagnetic wave signals (for example, millimeter-wave signals). The electromagnetic wave signals are reflected by various objects in the vehicle and then received by the radar's receiving antenna, thereby obtaining an echo signal reflected from the interior of the vehicle.

[0085] Step 204 : performing mean processing on the echo data in each frame of the echo data sequence to obtain an echo mean.

[0086] In implementations, each frame of echo signals contains a set of echo data sequences. The radar can then perform mean processing on the echo data in each frame of echo data sequence to obtain the echo mean corresponding to that frame of echo signals. Optionally, the echo mean can be the mean of the echo signal data within a fixed distance dimension interval, which is not limited in this embodiment of the present application.

[0087] Step 206 : Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each echo mean value to obtain an echo mean value data set.

[0088] In practice, after processing the echo mean data, the radar collects echo mean values equal to the number of sampling points M (i.e., a first predetermined number) within a preset fixed sampling window. A frame number is added to each echo mean value. The frame number indicates the sampling order of the echo signal corresponding to the echo mean value within the echo mean data set. Consequently, the radar obtains an echo mean data set. This echo mean data set can also be referred to as a feature matrix within the fixed sampling window.

[0089] For example, if set is used to represent the frame number, the frame numbers of the consecutive echo mean values in a corresponding sampling window are represented as: set1, set2, ... set M The echo mean is represented by amp, and the corresponding M echo mean values are represented as: amp1, amp2, ...amp M Furthermore, each echo mean value can be mapped to its frame number identifier to construct a feature matrix (represented by EnMatrix). The EnMatrix feature matrix contains matrix elements, each of which is a two-tuple containing the frame number identifier and the echo mean value corresponding to the frame number identifier. The feature matrix is specifically expressed as:

[0090] EnMatrix{{set1, amp1}, {set2, amp2},...{set M ,amp M}}

[0091] In one embodiment, Figure 3 As shown, the specific processing process of step 204 includes the following steps:

[0092] Step 302 : extracting envelope features contained in the echo mean value data set according to a preset first feature extraction algorithm, and performing statistical processing on the envelope features to determine a first feature quantity.

[0093] In implementation, the radar extracts features from each echo mean data set using a preset first feature extraction algorithm to obtain envelope features contained in the echo mean data set. The radar then counts all envelope features in the echo mean data set to obtain a first feature quantity representing the number of envelope features.

[0094] Specifically, the process of determining envelope features in the echo mean data set is as follows:

[0095] A peak value of echo mean data present in the echo mean data set is identified. For this peak value, if the values of the echo mean data for a third preset number of consecutive frame numbers (for example, the third preset number is number A, and number A is less than the first preset number M) increase in a gradient, and if the values of the echo mean data for a fourth preset number of consecutive frame numbers (for example, the fourth preset number is number B, and number B is less than the first preset number M, but B may be equal to or not equal to A) decrease in a gradient, then an envelope feature is determined in the echo mean data set. That is, the envelope feature includes the echo mean data and peak value data for the third preset number of consecutive frame numbers, and the echo mean data for the fourth preset number of consecutive frame numbers, and the third preset number of echo mean data and the fourth preset number of echo mean data satisfy a preset variation trend.

[0096] Optionally, when the radar extracts envelope features based on the first feature extraction algorithm, it can mark the echo mean data of the third preset number of consecutive frame numbers that are determined to be gradient rising as GradRiseUp is 1; using GradRiseUp as 1 as a priori judgment condition, on this basis, it is further judged whether there is a fourth preset number of consecutive frame numbers whose numerical values of the echo mean data are gradient decreasing.

[0097] Step 304 : extracting the burr features contained in the envelope features according to a preset second feature extraction algorithm, and performing statistical processing on the burr features to determine a second feature quantity.

[0098] In implementation, the radar can determine the burr feature in each envelope feature (i.e., the first feature) determined by each echo mean data set based on the extracted envelope feature according to the preset burr feature extraction algorithm. Then, the radar performs quantitative statistics on the burr features contained in each envelope feature to obtain the second feature quantity, which is expressed as burr.

[0099] Specifically, the process for determining a glitch feature is as follows: a first threshold (denoted by thr1) and a second threshold (denoted by thr2) are preset for the echo mean and the frame number identifier corresponding to the echo mean contained in each envelope feature. Then, to determine adjacent echo means, the radar calculates the interval difference between the frame number identifiers corresponding to any two echo means and the difference between the two echo means. If the calculated interval difference between the frame number identifiers corresponding to the two echo means is less than or equal to the first threshold, and the difference between the two echo means is less than or equal to the second threshold, then a glitch feature exists in the current envelope feature.

[0100] Optionally, the radar can count the number of burr features in each envelope feature based on the burr feature determination method and express it with burr. Then, the number of burr features in an echo mean data set (i.e., the second feature quantity) can be constructed as a burr matrix. For example, when an echo mean data set contains enNum envelope features, each of the enNum envelope features contains corresponding burr features. The number of burr features in each envelope feature (i.e., the second feature quantity) is counted. Then, based on the number of burr features corresponding to all envelope features in the echo mean data set, a burr matrix can be constructed. The burr matrix can be expressed as: EnBurrInforRe[Burr1, Burr2, Burr3,…,Burr enNum ]. Among them, Burr1 represents the number of burr features (second feature quantity) contained in the first envelope feature in the echo mean data set, and Burr2 represents the number of burr features contained in the second envelope feature in the echo mean data set. ... The meanings of the elements in the subsequent burr matrix are similar to them. Burr enNum It represents the number of burr features contained in the enNum-th envelope feature. The meaning of the elements in the burr matrix is not further described in this embodiment.

[0101] In one embodiment, Figure 4 As shown, the process of determining the envelope feature is described in detail, and the specific processing process of step 302 includes:

[0102] Step 402 : traverse each echo mean in the echo mean data set, and identify peak data in the echo mean data set.

[0103] In implementation, the echo mean data set includes EnMatrix {{set1, amp1}, {set2, amp2}, ... {set M ,amp M The radar traverses each echo mean in the echo mean data set, compares the echo mean values, identifies the peak data in the echo mean data set, and also determines the frame number corresponding to the peak data. For example, if there are enNum peak data in an echo mean data set, enNum echo mean values (peak data) and corresponding frame numbers can be obtained.

[0104] Step 404: for each peak data, obtain a second preset number of echo means including the peak data and the frame number identifier corresponding to the echo mean, and determine whether the echo mean and the frame number identifier corresponding to the echo mean are an envelope feature according to the preset envelope feature judgment condition.

[0105] The second preset number is equal to the sum of the third preset number, the fourth preset number and number one.

[0106] In implementation, for each peak data determined, the radar further determines whether the echo mean of this part meets the characteristic conditions of the preset envelope feature based on the peak data and the echo mean of the consecutive frames adjacent to the peak data. That is, if the numerical value of the echo mean data of the third preset number of consecutive frame numbers on the left side of the peak data increases gradiently, and the numerical value of the echo mean data of the fourth preset number of consecutive frame numbers on the right side of the peak data decreases gradiently, then the third preset number and fourth preset number of echo mean data, the peak data and the frame number corresponding to these echo mean data are determined to be an envelope feature.

[0107] Step 406 : Count the envelope features contained in the echo mean value data set to obtain a first feature value.

[0108] During implementation, the radar performs a statistical analysis of the envelope features contained in the mean echo data set to obtain a first characteristic quantity contained in the mean echo data set. For example, for the enNum peak data points, if each peak data point and the echo mean values of its left and right consecutive frames meet the envelope characteristic conditions, then the first characteristic quantity contained in the mean echo data set obtained by statistical analysis is enNum. The specific statistical analysis process is the same as that in step 104 above and is not further described in this embodiment.

[0109] In one embodiment, the radar pre-stores a plurality of target object detection and identification strategies and detection parameter thresholds corresponding to the detection and identification strategies of the target objects, such as a first feature value threshold and a second feature value threshold. After feature extraction is performed on each echo mean value data set, the radar detects the target object inside the target vehicle based on the detection and identification strategy and the extracted first feature value and second feature value, such as Figure 5 As shown, the specific processing process of step 206 includes:

[0110] Step 502 : When the first feature quantity is greater than a preset first feature quantity threshold, a target object detection result corresponding to the target vehicle to be detected is determined according to the second feature quantity and the preset second feature quantity threshold.

[0111] During implementation, the radar compares the first characteristic quantity (represented by enNum) with the preset first characteristic quantity threshold (represented by G) based on the detection and identification strategy. If the result that the first characteristic quantity is greater than the preset first characteristic quantity threshold, that is, enNum>G, it indicates that there is interference information inside the vehicle to be detected to determine whether there is a target object inside the vehicle. Then the radar needs to further determine the target object detection result corresponding to the target vehicle to be detected based on the second characteristic quantity and the preset second characteristic quantity threshold, that is, whether there is a target object inside the target vehicle to be detected. Among them, the second characteristic quantity threshold includes a first sub-threshold and a second sub-threshold. Specifically, the process of determining the target object detection result based on the second characteristic quantity and the preset second characteristic quantity threshold will be described in detail later and will not be repeated here. Step 504, when the first characteristic quantity is less than or equal to the first characteristic quantity threshold, determines that the target object detection result corresponding to the target vehicle to be detected is that there is a target object inside the vehicle.

[0112] In implementation, if the result obtained is that the first characteristic quantity is less than or equal to the preset first characteristic quantity threshold (represented by G), the radar determines that the target object detection result corresponding to the target vehicle to be detected is that there is a target object inside the vehicle, that is, there is a living object inside the vehicle.

[0113] In this embodiment, through the preset feature extraction algorithm and detection and recognition strategy, the discrimination processing of the first feature quantity and the second feature quantity extracted from the echo signal data set can effectively identify living objects and non-living objects in the vehicle, so as to avoid false alarms caused by the shaking of non-living objects in the vehicle and missed alarms of living objects.

[0114] In one embodiment, the second characteristic value threshold set inside the radar may include a first sub-threshold (denoted by thr3) and a second sub-threshold (denoted by thr4) of the second characteristic value, and the first sub-threshold is greater than the second sub-threshold, that is, thr3>thr4, as shown in FIG. Figure 6 As shown, when the first feature value is greater than the first feature value threshold, the specific processing process of step 502 includes:

[0115] Step 602: When all second feature quantities corresponding to the echo mean data set are less than or equal to the first sub-threshold, if all second feature quantities are greater than the second sub-threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that there is no target object inside the vehicle.

[0116] In implementation, the radar compares the second feature quantity with the first sub-threshold. If all the second feature quantities corresponding to the echo mean data set are less than or equal to the first sub-threshold, the radar further determines the difference between the second feature quantity and the preset second sub-threshold based on the second sub-threshold, and then determines the target object detection result corresponding to the target vehicle to be detected. Specifically, there are two cases:

[0117] Case 1: All the second feature quantities are greater than the second sub-threshold. This case indicates that the first feature quantity and the second feature quantity extracted from the echo mean data contained in the echo mean data set do not meet the conditions for the existence of the target object, that is, when the number of envelope features contained in the echo mean data set is greater than the first feature quantity threshold, and the number of burr features contained in each envelope feature in all the envelope features is greater than the second sub-threshold and less than or equal to the first sub-threshold, it is determined that there is no target object in the target vehicle to be detected.

[0118] The number of envelope features and the number of burr features can be used to characterize the frequency of motion. Specifically, when the envelope feature quantity (i.e., the first feature quantity) is less than or equal to the first feature quantity threshold, it is equivalent to detecting low-frequency motion. Specifically, it can be determined that there is a living object inside the target vehicle to be detected, and the living object produces low-frequency and large-amplitude motion. When the first feature quantity is greater than the first feature quantity threshold, the motion frequency is not determined at this time. Therefore, it is not possible to directly determine whether there is a living object. That is, the preliminary detection result may be subject to interference. In order to eliminate the influence of interference on the detection result, in this case, the burr features are further analyzed, and the burr feature analysis results are comprehensively analyzed to determine whether there is a living object. That is, the second feature quantity is compared with the second feature quantity threshold. In a specific implementation, the second feature quantity threshold includes a first sub-threshold and a second sub-threshold, and the first sub-threshold is greater than the second sub-threshold. When the second feature quantity is greater than the first sub-threshold, it is equivalent to detecting high-frequency motion. Specifically, it can be determined that there is a living object inside the target vehicle to be detected, and the living object produces high-frequency micro-motion. When the second feature value is less than the first sub-threshold and less than the second sub-threshold, it indicates that low-frequency motion of a living object has been detected. Specifically, it can be determined that a living object is present inside the target vehicle, producing low-frequency, large-amplitude motions. When the second feature value is less than the first sub-threshold but greater than the second sub-threshold, it indicates that an invalid target has been detected, meaning that no living object is present.

[0119] The radar then determines that the target object detection result corresponding to the target vehicle to be detected is that there is no target object inside the vehicle, that is, there is no living object inside the target vehicle to be detected.

[0120] In the second case, the second feature quantity is less than or equal to the second sub-threshold. This indicates that the echo mean data contained in the echo mean data dataset includes echo mean data generated when a target object undergoes low-frequency, large-amplitude motion. Consequently, the first and second feature quantities extracted from the echo mean data dataset satisfy the target object presence condition. The radar then determines that the target object detection result for the target vehicle is a target object inside the vehicle. This means that a living target is present inside the target vehicle. Based on the preset detection and identification strategy, this target object detection result can also be further subdivided into a target object exhibiting low-frequency, large-amplitude motion within the vehicle.

[0121] Step 604 : When there is a second characteristic value greater than the first sub-threshold in the echo mean value data set, it is determined that the target object detection result corresponding to the target vehicle to be detected is that there is a target object inside the vehicle.

[0122] During implementation, the radar compares all second feature quantities in the echo mean data set with the first sub-threshold. If any second feature quantity exceeds the first sub-threshold (i.e., thr3), it indicates that the echo mean data in the echo mean data set contains echo mean data generated by high-frequency micro-motions of the target object. Furthermore, if the first and second feature quantities extracted from the echo mean data set meet the target object presence condition, the radar determines that the target object detection result corresponding to the target vehicle is a target object inside the vehicle. Furthermore, based on a preset detection and identification strategy, this target object detection result can be further subdivided into a target object with high-frequency micro-motions inside the vehicle.

[0123] In one embodiment, an example of a target object detection method is provided, such as Figure 7 As shown in the figure, the specific processing process includes:

[0124] Step 701: Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo means.

[0125] Step 702: Determine a first feature quantity and a second feature quantity according to a preset feature extraction algorithm and an echo mean value data set.

[0126] Step 703, compare the first feature value (enNum) with the preset first feature value threshold (G). If the first feature value is greater than the preset first feature value threshold, execute step 704; if the first feature value is less than or equal to the preset first feature value threshold, execute step 707.

[0127] In step 704, all second feature quantities contained in the echo mean data set are compared with a preset second feature quantity threshold, where the second feature quantity threshold includes a first sub-threshold (thr3) and a second sub-threshold (thr4), and the first sub-threshold (thr3) is greater than the second sub-threshold (thr4). If all second feature quantities contained in the echo mean data set are less than or equal to the first sub-threshold (thr3), step 705 is executed; if any second feature quantity in the echo mean data set is greater than the first sub-threshold (thr3), step 707 is executed.

[0128] Step 705: If all second feature quantities in the echo mean value dataset are greater than the second sub-threshold (thr4), then execute step 706. If any second feature quantity in the echo mean value is less than or equal to the second sub-threshold (thr4), then execute step 707.

[0129] Step 706 : Determine that the target object detection result is that there is no target object inside the vehicle.

[0130] Step 707: Determine that the target object detection result indicates that the target object exists inside the vehicle.

[0131] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0132] Based on the same inventive concept, embodiments of the present application further provide a target object detection device for implementing the target object detection method involved above. The implementation solution provided by this device is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more target object detection device embodiments provided below can be referred to the above limitations of the target object detection method and will not be repeated here.

[0133] In one embodiment, Figure 8 As shown, a target object detection device 800 is provided, comprising: an acquisition module 810, a first determination module 820 and a second determination module 830, wherein:

[0134] An acquisition module 810 is configured to acquire an echo mean data set, wherein the echo mean data set includes a plurality of echo means, and the echo mean is determined by an echo signal reflected from the interior of the target vehicle to be detected;

[0135] A first determining module 820 is configured to perform feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features included in the echo mean data set, and the second feature quantity represents the number of burr features included in each envelope feature of the echo mean data set;

[0136] The second determination module 830 is configured to determine a target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and recognition strategy.

[0137] In one embodiment, the acquisition module 810 is configured to receive an echo signal reflected from the interior of a target vehicle to be detected, wherein each frame of the echo signal includes an echo data sequence;

[0138] Performing mean processing on the echo data in each frame of echo data sequence to obtain the echo mean;

[0139] Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each echo mean value to obtain an echo mean value data set.

[0140] In one embodiment, the first determining module 820 is configured to extract envelope features contained in the echo mean data set according to a preset first feature extraction algorithm and the echo mean data set, and perform statistical processing on the envelope features to determine a first feature quantity;

[0141] The burr feature contained in the envelope feature is extracted according to a preset second feature extraction algorithm, and the burr feature is statistically processed to determine a second feature quantity.

[0142] In one embodiment, the first determining module 810 is configured to traverse each echo mean in the echo mean data set and identify peak data in the echo mean data set;

[0143] For each peak data, obtaining a second preset number of echo means including the peak data and frame number identifiers corresponding to the echo means, and determining whether the echo means and the frame number identifiers corresponding to the echo means are an envelope feature according to a preset envelope feature discrimination condition;

[0144] Quantitative statistics are performed on the envelope features contained in the echo mean value data set to obtain a first feature quantity.

[0145] In one embodiment, the second determining module 830 is configured to determine a target object detection result corresponding to the target vehicle to be detected based on the second feature quantity and the preset second feature quantity threshold when the first feature quantity is greater than a preset first feature quantity threshold;

[0146] When the first feature quantity is less than or equal to the first feature quantity threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that the target object exists inside the vehicle.

[0147] In one embodiment, the second feature value threshold includes a first sub-threshold and a second sub-threshold of the second feature value, the first sub-threshold being greater than the second sub-threshold, and the second determination module 830 is configured to, when all corresponding second feature values in the echo mean value data set are less than or equal to the first sub-threshold, determine that the target object detection result corresponding to the target vehicle to be detected is that the target object does not exist inside the vehicle if all second feature values included in the echo mean value data set are greater than the second sub-threshold;

[0148] When there is a second feature value greater than the first sub-threshold in the echo mean value data set, it is determined that the target object detection result corresponding to the target vehicle to be detected is that the target object exists inside the vehicle.

[0149] Each module in the above-mentioned target object detection device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0150] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 9 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a target object detection method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse.

[0151] Those skilled in the art will understand that Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0152] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0153] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0154] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0155] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0156] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, and the like.

[0157] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0158] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A target object detection method, characterized in that: The method comprises: Acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected; performing feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features contained in the echo mean data set, and the second feature quantity represents the number of burr features contained in each envelope feature of the echo mean data set; Determining a target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and recognition strategy; The determining, based on the first feature quantity, the second feature quantity, and a preset detection and identification strategy, a target object detection result corresponding to the target vehicle to be detected includes: When the first feature quantity is greater than a preset first feature quantity threshold, determining a target object detection result corresponding to the target vehicle to be detected according to the second feature quantity and the preset second feature quantity threshold; When the first feature quantity is less than or equal to the first feature quantity threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

2. The method according to claim 1, characterized in that The step of obtaining an echo mean value data set includes: receiving an echo signal reflected from the interior of a target vehicle to be detected, wherein each frame of the echo signal includes an echo data sequence; performing mean processing on the echo data in each frame of the echo data sequence to obtain an echo mean; Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each of the echo mean values to obtain an echo mean value data set.

3. The method according to claim 1, characterized in that The step of extracting features from the echo mean value data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity includes: extracting envelope features contained in the echo mean value data set according to a preset first feature extraction algorithm, and performing statistical processing on the envelope features to determine a first feature quantity; The burr feature contained in the envelope feature is extracted according to a preset second feature extraction algorithm, and the burr feature is statistically processed to determine a second feature quantity.

4. The method according to claim 3, characterized in that The extracting the envelope features contained in the echo mean value data set according to a preset first feature extraction algorithm, and performing statistical processing on the envelope features to determine the first feature quantity includes: traversing each of the echo mean values in the echo mean value data set, and identifying peak data in the echo mean value data set; For each peak data, obtaining a second preset number of echo means including the peak data and frame number identifiers corresponding to the echo means, and determining whether the echo means and the frame number identifiers corresponding to the echo means are an envelope feature according to a preset envelope feature discrimination condition; Quantitative statistics are performed on the envelope features contained in the echo mean value data set to obtain a first feature quantity.

5. The method according to claim 1, wherein The second feature value threshold includes a first sub-threshold and a second sub-threshold of the second feature value, the first sub-threshold is greater than the second sub-threshold, and determining the target object detection result corresponding to the target vehicle to be detected based on the second feature value and the preset second feature value threshold includes: When all second feature quantities corresponding to the echo mean value dataset are less than or equal to the first sub-threshold, if all second feature quantities are greater than the second sub-threshold, determining that the target object detection result corresponding to the target vehicle to be detected is that there is no target object inside the vehicle; When a second feature value exists in the echo mean value data set and is greater than the first sub-threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

6. A target object detection device, characterized in that: The device comprises: an acquisition module, configured to acquire an echo mean value data set, wherein the echo mean value data set includes a plurality of echo mean values, and the echo mean values are determined by echo signals reflected from the interior of the target vehicle to be detected; a first determining module, configured to perform feature extraction on the echo mean data set according to a preset feature extraction algorithm to determine a first feature quantity and a second feature quantity; the first feature quantity represents the number of envelope features included in the echo mean data set, and the second feature quantity represents the number of burr features included in each envelope feature of the echo mean data set; A second determination module is configured to determine a target object detection result corresponding to the target vehicle to be detected based on the first feature quantity, the second feature quantity, and a preset detection and recognition strategy; The second determining module is specifically configured to determine a target object detection result corresponding to the target vehicle to be detected based on the second feature quantity and the preset second feature quantity threshold when the first feature quantity is greater than a preset first feature quantity threshold; When the first feature quantity is less than or equal to the first feature quantity threshold, it is determined that the target object detection result corresponding to the target vehicle to be detected is that a target object exists inside the vehicle.

7. The device according to claim 6, characterized in that The acquisition module is specifically configured to receive an echo signal reflected from the interior of the target vehicle to be detected, wherein each frame of the echo signal contains an echo data sequence; performing mean processing on the echo data in each frame of the echo data sequence to obtain an echo mean; Acquire echo mean values corresponding to a first preset number of consecutive frame echo signals, and add a frame number identifier to each of the echo mean values to obtain an echo mean value data set.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images