An object recognition method, apparatus, device, and storage medium

Abstract

This application discloses an object recognition method, apparatus, device, and storage medium. The method includes: acquiring a point cloud map generated by scanning objects within its own detection area using a Doppler sensor; determining a target area from the self-detection area based on the distribution information of points in the point cloud map; determining a target point set based on points in the target area whose frequency shift is greater than a first specific frequency shift; and determining a target object based on the target point set.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of intelligent control, and relate to, but are not limited to, an object recognition method and device, equipment, and a storage medium. BACKGROUND

[0002] A Doppler sensor generally transmits a signal through a transmitting antenna, receives a return signal formed by reflection of the signal by an object through a receiving antenna, and determines whether the object is a moving target by analyzing whether the return signal produces a frequency shift, so as to identify whether the object is a human body or an object. However, due to the complexity of an indoor environment, a wall surface and the like can reflect a signal, which easily causes a multipath phenomenon, that is, a virtual image of multiple objects is formed, thereby affecting object recognition. SUMMARY

[0003] In view of this, embodiments of the present application provide an object recognition method and device, equipment, and a storage medium.

[0004] In a first aspect, embodiments of the present application provide an object recognition method, which comprises: acquiring a point cloud map generated by scanning an object in a self-probe region by using a Doppler sensor; determining a target region from the self-probe region according to distribution information of points in the point cloud map; determining a target point set according to points in the target region whose frequency shift is greater than a first specific frequency shift; and determining a target object according to the target point set.

[0005] In a second aspect, embodiments of the present application provide an object recognition device, which comprises: an acquisition module configured to acquire a point cloud map generated by scanning an object in a self-probe region by using a Doppler sensor; a first determination module configured to determine a target region from the self-probe region according to distribution information of points in the point cloud map; a second determination module configured to determine a target point set according to points in the target region whose frequency shift is greater than a first specific frequency shift; and a third determination module configured to determine a target object according to the target point set.

[0006] In a third aspect, embodiments of the present application provide an object recognition equipment, which comprises a memory and a processor, and the memory stores a computer program capable of running on the processor, and the processor implements steps in the object recognition method of the first aspect of the present application when executing the computer program.

[0007] In a fourth aspect, embodiments of the present application provide a computer readable storage medium, which stores a computer program, and the computer program implements steps in the object recognition method of the first aspect of the present application when executed by a processor.

[0008] In the embodiment of the present application, the target region is determined according to the distribution information of the point cloud map generated by scanning the object by the Doppler sensor, and the target point set is determined according to the points in the target region whose frequency shift is greater than the first specific frequency shift, and then the target object is determined, so that only the points in the target region are analyzed to reduce the interference of the virtual image of the target object formed outside the target region due to the multipath effect on the judgment of the actual target object, and then the target object can be more accurately identified. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 A flowchart of an object identification method according to an embodiment of the present application is shown in

[0010] Figure 2 A schematic diagram of generating a point cloud map according to an embodiment of the present application is shown in

[0011] Figure 3 A schematic diagram of a method of curve fitting for points in a point cloud map according to an embodiment of the present application is shown in

[0012] Figure 4 A schematic diagram of a method of determining a target region according to an embodiment of the present application is shown in

[0013] Figure 5 A schematic diagram of another method of generating a point cloud map according to an embodiment of the present application is shown in

[0014] Figure 6 A schematic diagram of another method of curve fitting for points in a point cloud map according to an embodiment of the present application is shown in

[0015] Figure 7 A schematic diagram of another method of determining a target region according to an embodiment of the present application is shown in

[0016] Figure 8 A schematic diagram of the composition structure of an object identification device according to an embodiment of the present application is shown in

[0017] Figure 9 A schematic diagram of the hardware entity of an object identification device according to an embodiment of the present application is shown in DETAILED DESCRIPTION

[0018] The technical solutions of the present application are further described in detail below in combination with the drawings and embodiments.

[0019] Figure 1 A schematic diagram of the implementation flow of the object identification method according to an embodiment of the present application is shown in Figure 1 The method comprises:

[0020] Step 102: Obtain a point cloud map generated by scanning an object in a self-probing region by a Doppler sensor;

[0021] The Doppler sensor can be a sensor that measures the motion speed and other information of objects at different distances by using the principle of Doppler effect. It can measure the motion speed and other information of the object relative to the Doppler sensor by emitting microwave signals such as radar waves to the object, and then analyzing the frequency change of the echo signals reflected by the object, comparing the difference between the transmitted frequency and the reflected frequency, and the change of the transmitted phase and the reflected phase. The Doppler sensor can be a millimeter wave radar sensor.

[0022] The self-probing range can be a spatial range in which the millimeter wave radar sensor can receive echo signals reflected by the object. Since the probing range of the millimeter wave radar sensor is limited, some echo signals with a signal-to-noise ratio less than a specific signal-to-noise ratio threshold can not be received. Therefore, the self-probing range can be understood as a spatial range of echo signals with a signal-to-noise ratio not less than the signal-to-noise ratio threshold.

[0023] The object can be a wall, a cabinet, a table and chair, a person, a flowerpot, or any other thing that can be detected by the Doppler sensor. The point cloud map can be a map that can reflect the distribution of the object, formed by the echo signals reflected by the object surface irradiated by the signals emitted by the Doppler sensor. The echo signals carry the orientation and distance information of the object. Each object can generate a point set composed of multiple points when scanned by the Doppler sensor, and multiple objects can generate a point cloud map composed of multiple point sets. The distribution of points in the point set can reflect the size and shape of the object. Figure 2 A point cloud map generation schematic diagram of an embodiment of the present application is shown in Figure 2 The millimeter wave radar sensor 201 can generate a point cloud map 203 by scanning the objects in the self-probing range 202.

[0024] Step 104: determining a target region from the self-probing region according to the distribution information of the points in the point cloud map.

[0025] The distribution information of the points in the point cloud map can be used to determine the mutual relationship between the points in the point cloud map. The distribution information can be dense distribution, sparse distribution, or uniform distribution, etc. The target region can be a room region composed of a certain room, etc.

[0026] Due to different positions, different shapes, different reflectivities to radar waves, and other factors of different objects, the distribution of points in the generated point cloud map will show a certain regularity. For example, the reflectivity of a wall surface to radar waves is different from that of a chair. Therefore, the region where the wall surface is located can be determined according to the distribution information of the points in the point cloud map, and the room region can be determined from the self-probing region according to the region where the wall surface is located.

[0027] Step 106: determining a target point set according to points in the target region with frequency shift greater than a first specific frequency shift;

[0028] Wherein, the frequency shift, also known as Doppler shift, refers to the difference between the frequency of the transmitted signal and the echo signal caused by the Doppler effect due to the change of the distance between the Doppler sensor and the object. Since the frequency shift is positively correlated with the motion speed of the object, the point set composed of points with greater frequency shift can be determined from multiple point sets, and the motion speed of the object represented by these point sets is also relatively large.

[0029] Step 108: determining a target object according to the target point set.

[0030] Wherein, the target object can be a person. Since the person or animal object is moving, the wall, table, stool and other objects are generally stationary, so it can be considered that the target point set composed of points with greater frequency shift is likely to be a person or animal object, and the person can be identified in combination with the distribution law, shape, size and other factors of the points in the target point set.

[0031] In the embodiments of the present application, the target region is determined according to the distribution information of the point cloud generated by the Doppler sensor scanning the object, the target point set is determined according to the points in the target region with frequency shift greater than a first specific frequency shift, and then the target object is determined, so that only the points in the target region are analyzed to reduce the interference of the virtual image of the target object formed outside the target region due to the multipath effect on the judgment of the actual target object, and the target object can be identified more accurately.

[0032] The embodiments of the present application further provide an object method, which can include steps 202 to 212:

[0033] Step 202: acquiring a point cloud generated by scanning an object in a detection region of a Doppler sensor;

[0034] Step 204: determining the number of curves to be fitted according to the distribution information of the points in the point cloud;

[0035] Wherein, the fitting method can be curve fitting, which is a data processing method for approximating or imitating the functional relationship between the coordinates represented by a set of discrete points on a plane with a continuous curve. The curve fitting can be realized by the least square method.

[0036] Since the room is generally rectangular and the walls in the room are generally straight, the curve is generally a straight line, and at this time, the curve fitting can be linear fitting. Since the installation position of the Doppler sensor is different, the detection region of the Doppler sensor is different, and the distribution information of the points in the generated point cloud is also different, Figure 2The Doppler sensor can be installed on a wall in a room in a wall-mounted manner, and the Doppler sensor scans a point cloud generated by a detection area of the Doppler sensor. Figure 5 The Doppler sensor can be installed on a ceiling in a room in a ceiling-mounted manner, and the Doppler sensor scans a point cloud generated by a detection area of the Doppler sensor.

[0037] Referring to Figure 2 In a case where the Doppler sensor is a millimeter wave radar sensor 201, the detection area 202 is a sector, and the millimeter wave radar sensor 201 scans an object in the detection area 202 to generate a point cloud 203; in a case where the Doppler sensor is installed in a wall-mounted manner and the room is rectangular, three straight lines need to be fitted.

[0038] Referring to Figure 5 In a case where the Doppler sensor is a millimeter wave radar sensor (not shown in the figure), the detection area 501 is a circle, and the millimeter wave radar sensor scans an object in the detection area 501 to generate a point cloud 502; in a case where the Doppler sensor is installed in a ceiling-mounted manner and the room is rectangular, four straight lines need to be fitted.

[0039] Step 206: fitting a curve of the points in the detection area according to the distribution information of the points in the point cloud and the number of the lines, to obtain target curves of the lines after fitting;

[0040] Figure 3 and Figure 6 are schematic diagrams of methods of fitting curves of points in point clouds generated by Doppler sensors installed in wall-mounted and ceiling-mounted manners, respectively, referring to Figure 3 , three target curves, i.e., straight lines 301 to 303, are obtained after fitting; referring to Figure 6 , four target curves, i.e., straight lines 601 to 604, are obtained after fitting.

[0041] Step 208: determining a target area from the detection area according to a position of the target curve.

[0042] Figure 4 and Figure 7 are schematic diagrams of methods of determining target areas according to points in point clouds generated by Doppler sensors installed in wall-mounted and ceiling-mounted manners, respectively, referring to Figure 4According to the installation position of the Doppler sensor and the positions of the three straight lines, the position of the wall where the Doppler sensor is located and the positions of the other three walls can be determined, and then according to the positions of the four walls, the target region 401, i.e., the room region, can be determined. Points in the room region can be determined as valid point clouds, such as point 402, and points outside the room region can be determined as invalid point clouds, such as point 403; see Figure 7 According to the positions of the four straight lines, the positions of the four walls can be determined, and then the target region 701, i.e., the room region, can be determined. Points in the room region can be determined as valid point clouds, such as point 702, and points outside the room region can be determined as invalid point clouds, such as point 703.

[0043] Step 210: determining a target point set according to points in the target region with a frequency shift greater than a first specific frequency shift.

[0044] Step 212: determining a target object according to the target point set.

[0045] In steps 204 to 208 of the embodiments of the present application, the number of curves to be fitted is determined according to the distribution information of points in the point cloud map, and then the room region is determined according to the number and positions of the curves, so that the determination of the room region is more accurate.

[0046] The embodiments of the present application further provide an object recognition method, which can include steps 302 to 312.

[0047] Step 302: obtaining a point cloud map generated by scanning an object in a self-probing region with a Doppler sensor;

[0048] Step 304: determining a target region from the self-probing region according to the distribution information of points in the point cloud map;

[0049] Step 306: determining a set of points in the target region with a frequency shift greater than a first specific frequency shift as a set of points to be examined;

[0050] Step 308: screening out isolated points from the set of points to be examined by using a clustering algorithm;

[0051] The clustering algorithm can classify points in the set of points to be examined according to the similarity between points. The isolated points can be points that cannot be classified into a category with other points in the set of points to be examined. The clustering algorithm can be a K-MEANS algorithm (K-means clustering algorithm), a CLARANS algorithm (clustering algorithm based on random selection), a CHAMELEON algorithm (chameleon algorithm), a DENCLUE algorithm (clustering algorithm based on density), and a WAVE-CLUSTER algorithm (wavelet clustering algorithm), etc.

[0052] Step 310: determining a set of points except the isolated points in the point set to be considered as a target point set.

[0053] Step 312: determining a target object according to the target point set.

[0054] In steps 306 to 310 of the embodiment, by screening out the points with large frequency shift in the target region, i.e. screening out the points that are likely to constitute the target object (e.g. human body), and then screening out the outliers caused by noise from the points by the clustering algorithm, the determination of the target object can be more accurate.

[0055] The embodiment of the present application further provides an object recognition method, which can include steps 402 to 414.

[0056] Step 402: obtaining a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor;

[0057] Step 404: determining a target region from the self-probing region according to distribution information of points in the point cloud map;

[0058] Step 406: determining a target point set according to points with frequency shift greater than a first specific frequency shift in the target region;

[0059] Step 408: determining a target object according to the target point set;

[0060] Step 410: determining attribute information of the target point set;

[0061] The attribute information of the target point set can be frequency shift, position, etc. of points constituting the target point set, or can be the number of points constituting the target point set, etc.

[0062] Step 412: determining feature information of the target object according to the attribute information of the target point set;

[0063] Since each object can correspond to a point set composed of multiple points in the point cloud map, the target point set can be divided into different point sets corresponding to different objects according to the frequency shift, position and distribution information, etc. of points in the target point set, e.g. the target point set can be divided into a first point set corresponding to a first object and a second point set corresponding to a second object according to the frequency shift and position, etc. of points, and the feature information of the first object is determined according to the attribute information of the first point set, and the feature information of the second object is determined according to the attribute information of the second point set; the feature information of the target object includes at least one of the following: frequency shift of the target object, position of the target object, and number of the target object.

[0064] Step 414: controlling the device according to the characteristic information of the target object.

[0065] The device can be a computer, a television, an air conditioner, a smart fan, a smart lamp, etc., and the Doppler sensor can be installed on the device.

[0066] In steps 410-414 of the embodiment, the device is controlled according to the characteristic information of the target object determined according to the attribute information of the target point set, which can improve the intelligence of the device and the pertinence of the device to the target object; and the diversity of the characteristic information of the target object determines the flexibility and diversity of the device control mode.

[0067] The embodiment further provides an object recognition method, which can include steps 502-516.

[0068] Step 502: obtaining a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor;

[0069] Step 504: determining a target region from the self-probing region according to distribution information of points in the point cloud map;

[0070] Step 506: determining a target point set according to points in the target region with a frequency shift greater than a first specific frequency shift;

[0071] Step 508: determining a target object according to the target point set;

[0072] Step 510: determining attribute information of the target point set;

[0073] Step 512: determining characteristic information of the target object according to the attribute information of the target point set;

[0074] Step 514: in the case where the characteristic information of the target object is a frequency shift of the target object and a position of the target object, determining a working mode of a device according to the frequency shift of the target object and the position of the target object;

[0075] The working mode of the device can be different according to different devices; in the case where the device is an air conditioner, the working mode of the device can be a blowing mode and / or a wind speed of the air conditioner; in the case where the device is a lamp, the working mode of the device can be a control mode of the lamp, which can be used to control the brightness and color of the lamp; in the case where the device is a television, the working mode of the device can be a television mode, a VOD (Video on Demand) mode and an application mode of the television.

[0076] Step 516: controlling the device to work in the working mode.

[0077] In one embodiment, according to the frequency shift, position and distribution information of points constituting the target object of the target object, it can be determined that the target object is a child, and in the case that the device is a television, the television can be controlled to work in a television mode, so as to avoid the child from accessing the network too early through the VOD mode and other applications, and prevent the child from being addicted to the network.

[0078] In one embodiment, according to the frequency shift, position and distribution information of points constituting the target object of the target object, it can be determined that the target object is a person, and the frequency shift of the person is in a specific frequency shift range, it can be judged that the person is writing at a desk, and in the case that the device is a lamp, the light of the lamp can be controlled to keep in an eye protection mode, so as to play a role in protecting the vision of the person.

[0079] In the embodiments of the present application, the working mode of the device is determined according to the frequency shift and position of the target object, so that the determined working mode of the device is more targeted for the target object.

[0080] The embodiments of the present application further provide an object recognition method, which can include steps 602 to 618.

[0081] Step 602: acquiring a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor;

[0082] Step 604: determining a target region from the self-probing region according to distribution information of points in the point cloud map;

[0083] Step 606: determining a target point set according to points with a frequency shift greater than a first specific frequency shift in the target region;

[0084] Step 608: determining a target object according to the target point set;

[0085] Step 610: determining attribute information of the target point set;

[0086] Step 612: determining feature information of the target object according to the attribute information of the target point set;

[0087] Step 614: in the case that the feature information of the target object is a frequency shift of the target object and a position of the target object, the device is an air conditioner, the working mode of the device includes a wind speed and a blowing mode of the air conditioner, and the frequency shift of the target object is greater than a second specific frequency shift, determining that the wind speed of the air conditioner is a first wind speed, and determining that the blowing mode of the air conditioner is a first blowing mode;

[0088] Step 616: determining the air speed of the air conditioner as a second air speed and determining the air supply mode of the air conditioner as a second air supply mode in a case where the frequency shift of the target object is not greater than the second specific frequency shift;

[0089] wherein the second specific frequency shift is greater than the first specific frequency shift; the first air speed is greater than the second air speed; the first air supply mode is a mode of air supply towards the position of the target object; and the second air supply mode is a mode of air supply away from the position of the target object.

[0090] wherein, in a case where the target object is a person, the movement speed of the person can be determined according to the size of the frequency shift; since the person needs to be cooled as soon as possible in an active state, the air conditioner needs to blow towards the person at a relatively high air speed; and in a quiet state (such as a sleep state), the air conditioner needs to blow away from the person at a relatively low air speed; therefore, the air supply mode and the air speed of the air conditioner can be determined according to the movement speed and the position of the person.

[0091] Step 618: controlling the device to work in the working mode.

[0092] In the embodiments of the present application, the air speed and the air supply mode of the air conditioner are determined according to the size of the frequency shift of the target object, so that the intelligence of air supply of the air conditioner can be improved.

[0093] The embodiments of the present application further provide an object recognition method, which can include steps 702 to 716:

[0094] Step 702: acquiring a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor;

[0095] Step 704: determining a target region from the self-probing region according to distribution information of points in the point cloud map;

[0096] Step 706: determining a target point set according to points with a frequency shift greater than a first specific frequency shift in the target region;

[0097] Step 708: determining a target object according to the target point set;

[0098] Step 710: determining attribute information of the target point set;

[0099] Step 712: determining feature information of the target object according to the attribute information of the target point set;

[0100] Step 714: determining a target position of a device to be worked according to the position of the target object in a case where the feature information of the target object is the position of the target object;

[0101] When the devices are multiple, distances between the multiple devices and the target object can be determined, and a target position of a device closest to the target object is determined, and the device at the target position is in an operating state.

[0102] Step 716: controlling the device at the target position to operate.

[0103] In the embodiments of the present application, the target position of the device that needs to operate is determined according to the position of the target object, so that the device close to the target object can be controlled to operate, and the purpose of saving resources is achieved.

[0104] The embodiments of the present application further provide an object recognition method, which can include steps 802 to 816.

[0105] Step 802: obtaining a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor;

[0106] Step 804: determining a target region from the self-probing region according to distribution information of points in the point cloud map;

[0107] Step 806: determining a target point set according to points with a frequency shift greater than a first specific frequency shift in the target region;

[0108] Step 808: determining a target object according to the target point set;

[0109] Step 810: determining attribute information of the target point set;

[0110] Step 812: determining feature information of the target object according to the attribute information of the target point set;

[0111] Step 814: when the feature information of the target object is a number of the target objects, determining a target number of devices that need to operate according to the number of the target objects;

[0112] When the devices are multiple, the target number of devices that need to operate can be determined according to the number of the target objects, and the devices at the target number are in an operating state; in an embodiment, when the target object is a person, the number of the persons is 6, and the device is a lamp, it can be determined that the number of the persons is large, and multiple lamps need to be turned on, for example, 3 lamps are turned on.

[0113] Step 816: controlling the devices at the target number to operate.

[0114] In the embodiment of the present application, the target number of devices that need to work is determined according to the number of target objects, so that the target number of devices can be controlled to work, thereby achieving the purpose of saving resources.

[0115] In the embodiment of the present application, an air conditioner is taken as an example to illustrate the object recognition device. In the implementation process, the air conditioner and the object recognition device can be two independent devices, so that the object recognition device interacts with the air conditioner. Of course, the air conditioner and the object recognition device can be the same device as in the embodiment. In this way, the air conditioner can obtain a point cloud map generated by the millimeter wave radar sensor mounted on the air conditioner when scanning an object in a detection area of the air conditioner, and determine a target object according to distribution information of the point cloud map, and control the air conditioner according to feature information of the target object. In the case where the target object is a user, the air conditioner currently mounted with the millimeter wave radar sensor detects feature information such as the position, behavior, and vital signs of the user to determine the air conditioning function required by the user, and realizes automatic and intelligent operation of the air conditioner.

[0116] The millimeter wave radar sensor currently used in the air conditioner transmits millimeter waves through a transmitting antenna, receives a return wave through a receiving antenna, calculates the distance of a target through the time difference of the received return wave, and judges whether it is a moving target by analyzing whether the return wave produces frequency shift (Doppler effect), so as to distinguish between a human body and a stationary object, and thus realize intelligent air supply and other intelligent control of the air conditioner according to the position of the human body. However, due to the complex indoor environment, the wall surface can reflect the millimeter wave, which is easy to cause the so-called multipath phenomenon, that is, after the radar wave emitted is reflected by the human body, part of it is directly returned to the radar sensor, but part of it is reflected to the wall, and then received by the receiving antenna. Or another situation is that the radar wave is reflected to the human body by the wall and then emitted back to the radar module. Such multipath phenomenon is easy to cause false detection, that is, when a person moves in a certain place, the radar module will also detect a moving target in another direction or directions, which is easy to cause false detection.

[0117] The embodiment of the present application discloses a room size detection method based on millimeter wave radar detection, a human body detection method based on room size, and an air conditioner intelligent control method and system. The position of each wall in the room is obtained by least square estimation of all return wave signals of the millimeter wave radar, and the room size is obtained. All possible human bodies and their positions are detected by Doppler effect, and the targets beyond the room are filtered out according to the room size, and the remaining moving targets are determined as human bodies. The air conditioner controls the intelligent air supply according to the position of the human body. When the wind blows on the human body, the air deflector is adjusted to the direction of the human body, so that the wind blows to the position of the human body. When the wind avoids the human body, the air deflector is adjusted to the direction without the human body, so that the wind blows away from the position of the human body.

[0118] For the air conditioner in the wall-mounted installation mode, such as Figure 2As shown, the millimeter-wave radar 201 emits radar waves horizontally. All objects reflect these radar waves, and different materials have different reflectivities. Walls generally have higher reflectivity. The receiving antenna receives all echo signals. Based on the signal-to-noise ratio of the echo signals exceeding a preset threshold, a point cloud map 203 is obtained within the detection range 202. It is generally assumed that the walls of a room are flat; therefore, for wall-mounted installations, reflections from three walls can be considered. The least squares method can be used to estimate the reflections. Figure 3 The lines 301 to 303 are shown.

[0119] For ceiling-mounted air conditioners, such as Figure 5 As shown, the millimeter-wave radar emits radar waves vertically from top to bottom, and the receiving antenna receives all the echo signals. Based on the signal-to-noise ratio of the echo signals exceeding a preset threshold, a point cloud map 502 is obtained within the detection range 501. Similarly, assuming the room walls are flat, for ceiling-mounted installations, reflections from all four walls can be considered. The least squares method can be used to estimate the reflections as shown below. Figure 6 The lines 601 to 604 are shown.

[0120] After determining the room size, see Figure 4 Only point clouds within room 401 are considered valid point clouds, such as point cloud 402; similarly, see [reference needed]. Figure 7 Only point clouds within room 701 are considered valid point clouds, such as point cloud 702; at the same time, a clustering algorithm is used to further eliminate noise interference, thereby detecting human bodies.

[0121] This application discloses a room size detection method based on millimeter-wave radar, a human body detection method based on room size, and an intelligent air conditioning control method and system. By analyzing the radar echo signal, the room size is estimated, and false detections of human bodies caused by multipath propagation are eliminated based on the room size, thereby improving the accuracy of indoor human body detection using millimeter-wave radar.

[0122] Based on the foregoing embodiments, this application provides an object recognition device, which includes various units and modules included in each unit. It can be implemented by a processor in the object recognition device; of course, it can also be implemented by specific logic circuits. In the implementation process, the processor can be a central processing unit (CPU), a microprocessor unit (MPU), a digital signal processor (DSP), or a field programmable gate array (FPGA), etc.

[0123] Figure 8As shown in the constituent structure diagram of the object recognition device of the embodiments of the present application, Figure 8 The object recognition device comprises an acquisition module 801, a first determination module 802, a second determination module 803 and a third determination module 804, wherein:

[0124] The acquisition module 801 is configured to acquire a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor.

[0125] The first determination module 802 is configured to determine a target region from the self-probing region according to distribution information of points in the point cloud map.

[0126] The second determination module 803 is configured to determine a target point set according to points in the target region with a frequency shift greater than a first specific frequency shift.

[0127] The third determination module 804 is configured to determine a target object according to the target point set.

[0128] In some embodiments, the first determination module 802 comprises: a first determination unit configured to determine a number of curves to be fitted according to the distribution information of points in the point cloud map; a fitting unit configured to perform curve fitting on points in the self-probing region according to the distribution information of points in the point cloud map and the number to obtain target curves of the number after fitting; and a second determination unit configured to determine a target region from the self-probing region according to a position of the target curves.

[0129] In some embodiments, the second determination module 803 comprises: a third determination unit configured to determine a set of points in the target region with a frequency shift greater than a first specific frequency shift as a set of points to be investigated; a screening unit configured to screen out isolated points from the set of points to be investigated by using a clustering algorithm; and a fourth determination unit configured to determine a set of points in the set of points to be investigated except the isolated points as a target point set.

[0130] In some embodiments, the device further comprises: a fourth determination module configured to determine attribute information of the target point set; a fifth determination module configured to determine feature information of the target object according to the attribute information of the target point set; and a control module configured to control a device according to the feature information of the target object.

[0131] In some embodiments, the feature information of the target object comprises at least one of a frequency shift of the target object, a position of the target object and a number of the target object.

[0132] In some embodiments, when the feature information of the target object is the frequency shift of the target object and the position of the target object, the control module comprises: a fifth determination unit configured to determine the working mode of the device according to the frequency shift of the target object and the position of the target object; and a first control unit configured to control the device to work in the working mode.

[0133] In some embodiments, the device comprises an air conditioner, and when the device is an air conditioner, the working mode of the device comprises the air speed and the air supply mode of the air conditioner, and the fifth determination unit comprises: a first determination sub-unit configured to determine the air speed of the air conditioner as a first air speed and the air supply mode of the air conditioner as a first air supply mode when the frequency shift of the target object is greater than a second specific frequency shift; and a second determination sub-unit configured to determine the air speed of the air conditioner as a second air speed and the air supply mode of the air conditioner as a second air supply mode when the frequency shift of the target object is not greater than the second specific frequency shift; wherein the second specific frequency shift is greater than the first specific frequency shift; the first air speed is greater than the second air speed; the first air supply mode is a mode of air supply towards the position of the target object; and the second air supply mode is a mode of air supply away from the position of the target object.

[0134] In some embodiments, when the feature information of the target object is the position of the target object, the control module comprises: a sixth determination unit configured to determine the target position of the device that needs to work according to the position of the target object; and a second control unit configured to control the device corresponding to the target position to work.

[0135] In some embodiments, when the feature information of the target object is the number of the target objects, the control module comprises: a seventh determination unit configured to determine the target number of the device that needs to work according to the number of the target objects; and a third control unit configured to control the device of the target number to work.

[0136] The above description of the device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the device embodiments of the present application, please refer to the description of the method embodiments of the present application for understanding.

[0137] It should be noted that, in the embodiments of the present application, if the object recognition method described above is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, and includes several instructions for causing an object recognition device to execute all or part of the methods described in the embodiments of the present application. The storage medium mentioned above includes: a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk, and various media that can store program codes. Thus, the embodiments of the present application are not limited to any specific hardware and software combination.

[0138] Correspondingly, the embodiments of the present application provide an object recognition device, Figure 9 A hardware entity diagram of the object recognition device of the embodiments of the present application is shown in Figure 9 The hardware entity of the object recognition device 900 includes a memory 901 and a processor 902, the memory 901 stores a computer program executable on the processor 902, and the processor 902 executes the computer program to realize the steps of the object recognition method provided in the above embodiments.

[0139] The memory 901 is configured to store instructions and applications executable by the processor 902, and can also cache data to be processed by the processor 902 and each module in the object recognition device 500 (for example, image data, audio data, voice communication data and video communication data), which can be realized by a flash (FLASH) or a random access memory (RAM).

[0140] Correspondingly, the embodiments of the present application provide a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the steps of the object recognition method provided in the above embodiments.

[0141] It should be noted that: the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects to the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present application, please refer to the description of the method embodiments of the present application for understanding.

[0142] It should be understood that every feature, structure, or characteristic described herein is within a preferred embodiment of the present application. It should be noted that the foregoing embodiments are merely exemplary and are not to be construed as limiting the present application. It should also be noted that features described in the foregoing relate to both structural and method aspects of the application. Accordingly, the terminology in use has a broad meaning from the context of use.

[0143] It should be noted that, as used in this document, the terms "comprises" or "comprising," or "includes" or "including" or "has" or "having" or "contains" or "containing" or variants thereof, are intended to cover a non-exclusive inclusion. For example, a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Further, unless otherwise specified, terms such as "first" and "second" are used merely as labels, and are not intended to impose numerical requirements on their objects.

[0144] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The above-described device embodiments are merely illustrative, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed components can be through some interfaces, indirect coupling or communication connection between devices or units, which can be electrical, mechanical or other forms.

[0145] The units described above as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units; they can be located in one place or distributed on multiple network units; some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment. In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be realized in the form of hardware or in the form of hardware plus software functional unit.

[0146] Those skilled in the art can understand that all or part of the steps of the foregoing method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program performs the steps of the foregoing method embodiments when executed; and the foregoing storage medium includes a mobile storage device, a read only memory (Read Only Memory, ROM), a magnetic disc or an optical disc, and various media that can store program codes. Alternatively, the integrated units of the present application can be stored in a computer readable storage medium if they are realized in the form of software function modules and sold or used as independent products. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, includes a plurality of instructions for causing an object recognition device to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes a mobile storage device, a ROM, a magnetic disc or an optical disc, and various media that can store program codes.

[0147] The methods disclosed in the several method embodiments provided by the present application can be combined arbitrarily without conflict to obtain new method embodiments. The features disclosed in the several product embodiments provided by the present application can be combined arbitrarily without conflict to obtain new product embodiments. The features disclosed in the several method or device embodiments provided by the present application can be combined arbitrarily without conflict to obtain new method or device embodiments.

[0148] The above is only an implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of object recognition, characterized by, The method comprises: acquiring a point cloud map generated by scanning an object in a self-detection area by using a Doppler sensor; determining a target area from the self-detection area according to distribution information of points in the point cloud map, and determining points outside the target area as invalid point clouds; the target area is a room area composed of a room in which the Doppler sensor is located; the room area is determined by a wall area obtained by processing the distribution information of points in the point cloud map; determining a target point set according to points in the target area with a frequency shift greater than a first specific frequency shift; determining a target object according to the target point set; the determining of the target area from the self-detection area according to the distribution information of points in the point cloud map comprises: determining the number of curves to be fitted according to the distribution information of points in the point cloud map; performing curve fitting on points in the self-detection area according to the distribution information of points in the point cloud map and the number, to obtain target curves of the number after fitting; determining the target area from the self-detection area according to the position of the target curves.

2. The method of claim 1, wherein, the determining of the target point set according to points in the target area with a frequency shift greater than a first specific frequency shift comprises: determining a set of points in the target area with a frequency shift greater than a first specific frequency shift as a set of points to be investigated; screening out isolated points from the set of points to be investigated by using a clustering algorithm; determining a set of points in the set of points to be investigated except the isolated points as the target point set.

3. The method of claim 1, wherein, The method further comprises: determining attribute information of the target point set; determining feature information of the target object according to the attribute information of the target point set; controlling a device according to the feature information of the target object.

4. The method of claim 3, wherein, In the case that the feature information of the target object is a frequency shift of the target object and a position of the target object, the controlling of the device according to the feature information of the target object comprises: determining a working mode of the device according to the frequency shift of the target object and the position of the target object; controlling the device to work in the working mode.

5. The method of claim 4, wherein, The device comprises an air conditioner; in the case that the device is an air conditioner, the working mode of the device comprises a wind speed and a blowing mode of the air conditioner, and the determining of the working mode of the device according to the frequency shift of the target object and the position of the target object comprises: in the case that the frequency shift of the target object is greater than a second specific frequency shift, determining the wind speed of the air conditioner as a first wind speed and determining the blowing mode of the air conditioner as a first blowing mode; in the case that the frequency shift of the target object is not greater than the second specific frequency shift, determining the wind speed of the air conditioner as a second wind speed and determining the blowing mode of the air conditioner as a second blowing mode; wherein the second specific frequency shift is greater than the first specific frequency shift; the first wind speed is greater than the second wind speed; the first blowing mode is a mode of blowing towards the position of the target object; and the second blowing mode is a mode of blowing away from the position of the target object.

6. The method of claim 3, wherein, In a case where the characteristic information of the target object is a position of the target object, the controlling the device according to the characteristic information of the target object comprises: determining a target position of the device to be operated according to the position of the target object; and controlling the device corresponding to the target position to operate.

7. The method of claim 3, wherein, In a case where the characteristic information of the target object is a quantity of the target object, the controlling the device according to the characteristic information of the target object comprises: determining a target quantity of the device to be operated according to the quantity of the target object; and controlling the device of the target quantity to operate.

8. An object recognition apparatus characterized by comprising: The apparatus comprises: an acquisition module configured to acquire a point cloud map generated by scanning an object in a self-probing region by using a Doppler sensor; a first determination module configured to determine a target region from the self-probing region according to distribution information of points in the point cloud map, and determine points outside the target region as invalid points; the target region is a room region composed of a room in which the Doppler sensor is located; the room region is determined by processing a region in which a wall surface is located according to the distribution information of points in the point cloud map; a second determination module configured to determine a target point set according to points in the target region with a frequency shift greater than a first specific frequency shift; a third determination module configured to determine a target object according to the target point set. The first determination module comprises: a first determination unit configured to determine a number of curves to be fitted according to the distribution information of points in the point cloud map; and a fitting unit configured to perform curve fitting on points in the self-probing region according to the distribution information of points in the point cloud map and the number, to obtain target curves of the number after fitting; a second determination unit configured to determine a target region from the self-probing region according to a position of the target curves.

9. An object recognition device comprising a memory and a processor, said memory storing a computer program operable on the processor, characterized in that, The processor implements the steps in the object recognition method in any one of claims 1 to 7 when executing the program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the steps in the object recognition method in any one of claims 1 to 7 when executed by the processor.

Images