WO2018121330A1 - 一种检测方法、相关装置和系统 - Google Patents

一种检测方法、相关装置和系统 Download PDF

Info

Publication number
WO2018121330A1
WO2018121330A1 PCT/CN2017/117065 CN2017117065W WO2018121330A1 WO 2018121330 A1 WO2018121330 A1 WO 2018121330A1 CN 2017117065 W CN2017117065 W CN 2017117065W WO 2018121330 A1 WO2018121330 A1 WO 2018121330A1
Authority
WO
WIPO (PCT)
Prior art keywords
detection
result
data
determining
movable body
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/117065
Other languages
English (en)
French (fr)
Inventor
周明春
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alibaba Group Holding Ltd
Original Assignee
Alibaba Group Holding Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Alibaba Group Holding Ltd filed Critical Alibaba Group Holding Ltd
Publication of WO2018121330A1 publication Critical patent/WO2018121330A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01VGEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V11/00Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00

Definitions

  • the present invention relates to the field of data processing, and in particular to a detection method, related apparatus and system.
  • detection technology it has been possible to apply detection techniques to device control in buildings, such as the common control of open voice control lamps by detecting sound.
  • detection of a movable body may be included, and the movable body may include a person, an animal, a movable robot, or the like.
  • the detection result of the movable body can be applied to the control of various devices in the building. For example, when it is detected that there is no movable body in a conference room, the system can turn off the air conditioning lighting and the like in the conference room to save energy. . It can be seen that the detection accuracy of the movable body is of great significance for the automatic control of the building.
  • the traditional movable body detection methods mainly include single detection methods such as infrared remote sensing detection and ultrasonic radar detection. Any one of the detection methods has its own limitations, so the accuracy of this single detection method is not high, and often errors occur.
  • the room of the original person is judged to be unmanned, causing the system to turn off the room lighting and air conditioning in the case of an actual person, and a bad user experience is generated for the person in the room.
  • the present invention provides a detection method, a related device and a system, which can reduce the adverse effect of a single type of detection device on the user experience due to technical limitations due to technical limitations, and effectively improve detection.
  • the accuracy is a measure of the accuracy of the detection device.
  • the present invention provides a detection method, which is applied to a detection server, wherein the detection server is configured with a detection model, and the detection model has a determination weight corresponding to different types of detection devices, and the method includes:
  • the detection model obtains an output parameter according to the determination weight corresponding to the type of the plurality of detection devices and the plurality of input parameters;
  • a determination result as to whether there is a movable body in the specified area is obtained according to the output parameter.
  • the method further includes:
  • the determination weight corresponding to the detection device type of the detection result identical to the actual result in the detection model is correspondingly increased.
  • the detection model has multiple data sequences, each data sequence includes M data bits, each data bit corresponding to one type of detection device, and the corresponding increase in the detection model and the actual result
  • the determining weights corresponding to the types of detecting devices of the same detection result include:
  • the target sequence is a data sequence, and the target sequence outputs the same detection as the actual result.
  • the result of the detection device is filled with a valid factor in the data bit, and the other data bits are filled with an invalid factor;
  • Adjusting the total number of valid factors filled in the data bits of all the data sequences of the detection model and the data bits corresponding to the detection devices outputting the same detection result as the actual result in all data sequences Determining a weight corresponding to the detection device type of the detection result that is the same as the actual result; wherein, for the first device, the total number of valid factors filled in all data sequences according to the detection model, and the first in all data sequences Determining, by the total number of valid factors filled in the data bits corresponding to the device, the determining weight corresponding to the first device, where the first device is detecting any one of the detecting devices that output the detection result with the same actual result device.
  • the determining, according to the output parameter, whether the mobile device has a movable body in the specified area includes:
  • the first data is used to identify data having a movable body
  • the second data is used to identify that there is no movable Volume data
  • the type of the detecting device includes at least a human body occupancy sensor, a positioning sensor, and a noise sensor.
  • the plurality of detection results are specifically N
  • the plurality of input parameters are specifically N
  • the determining, by the plurality of detection results, a plurality of input parameters are input into the detection model include:
  • the N detection results are input to the detection model as N input parameters.
  • the types of the N detecting devices are different.
  • the multiple detection results are specifically N, the multiple input parameters are specifically M, M is less than N, and the N detecting devices have M different types, and the multiple detections are performed.
  • the result determines that a plurality of input parameters are input into the detection model, including:
  • the M detection results are input to the detection model as M input parameters.
  • the present invention provides a detection method, which is applied to an interaction end, and the method includes:
  • an adjustment instruction is output, and the adjustment instruction is used to indicate that the determination result is erroneous.
  • the method further includes: if the determination result is the same as whether the actual result of the movable body in the designated area is the same, the controllable device in the designated area is controlled according to the determination result.
  • controllable device comprises a combination of any one or more of a lighting device, an air conditioner, a display device, and a projection device.
  • the present invention provides a detection server, wherein the detection server is configured with a detection model, wherein the detection model has a determination weight corresponding to different types of detection devices, and the detection server includes an acquisition unit and a determination unit. :
  • the acquiring unit is configured to acquire, by the plurality of detecting devices, a plurality of detection results for the movable area in the specified area, and each of the plurality of detecting devices outputs a detection result ;
  • the determining unit is configured to determine, from the plurality of detection results, that a plurality of input parameters are input into the detection model;
  • the determining unit is further configured to use the detection model to obtain an output parameter according to the determining weight corresponding to the type of the plurality of detecting devices and the plurality of input parameters;
  • the determining unit is further configured to obtain, according to the output parameter, a determination result of whether there is a movable body in the designated area.
  • the detection server further includes a determining unit and an adding unit:
  • the determining unit is configured to determine whether the plurality of detection results are the same as the actual result. ;
  • the adding unit is triggered, and the adding unit is configured to correspondingly increase the determining weight corresponding to the detecting device type of the detection result that is the same as the actual result in the detection model.
  • the detection model has a plurality of data sequences, each of the data sequences includes M data bits, each data bit corresponding to one type of detecting device, and the adding unit is further configured to: according to the determining Whether the detection result is the same as the actual result in the detection result, the target sequence is established, and the target sequence is a data sequence, and the target device outputs the corresponding detection device corresponding to the actual result.
  • the data bits are filled with valid factors, and other data bits are filled with invalid factors;
  • Adjusting the total number of valid factors filled in the data bits of all the data sequences of the detection model and the data bits corresponding to the detection devices outputting the same detection result as the actual result in all data sequences Determining a weight corresponding to the detection device type of the detection result that is the same as the actual result; wherein, for the first device, the total number of valid factors filled in all data sequences according to the detection model, and the first in all data sequences Determining, by the total number of valid factors filled in the data bits corresponding to the device, the determining weight corresponding to the first device, where the first device is detecting any one of the detecting devices that output the detection result with the same actual result device.
  • the determining unit includes a determining subunit and determining the subunit:
  • the determining subunit is configured to determine a degree of proximity between the output data of the detection model and the first data and the second data, where the first data is used to identify data having a movable body, and the second The data is data for identifying that there is no movable body;
  • the determining subunit is triggered, and the determining subunit is configured to obtain a determination result that the movable area is in the designated area;
  • the determining subunit is triggered, and the determining subunit is configured to obtain a determination result that there is no movable body in the designated area.
  • the type of the detecting device includes at least a human body occupancy sensor, a positioning sensor, and a noise sensor.
  • the multiple detection results are specifically N
  • the multiple input parameters are specifically N
  • the determining unit is further configured to input the N detection results as the N input parameters to the detection.
  • the model In the model.
  • the types of the N detecting devices are different.
  • the multiple detection results are specifically N
  • the multiple input parameters are specifically M
  • M is smaller than N
  • the N detecting devices have M different types
  • the determining unit is further used to The detection results corresponding to the detection devices of this type are determined by the detection results of the X detection devices of one type, so that M detection results are determined from the N detection results, and X is greater than Y.
  • X is less than N
  • Y is less than M;
  • the M detection results are input to the detection model as M input parameters.
  • the present invention provides an interaction end, where the interaction end includes an acquisition unit and an output unit:
  • the obtaining unit is configured to acquire a determination result of whether there is a movable body in the designated area, where the specified area has multiple detecting devices, and each of the plurality of detecting devices having multiple different types is detected
  • the device outputs a detection result, which is determined according to the plurality of detection results of the plurality of detection devices for whether there is a movable body in the designated area;
  • the output unit is triggered to output an adjustment instruction, where the adjustment instruction is used to indicate that the determination result is wrong.
  • control unit also includes a control unit:
  • the control unit is triggered, and the control unit is configured to control the controllable device in the designated area according to the determination result.
  • controllable device comprises a combination of any one or more of a lighting device, an air conditioner, a display device, and a projection device.
  • the present invention provides a detection system including a detection server and an interaction end:
  • a detection model is configured in the detection server, and the detection model has a determination weight corresponding to different types of detection devices, and the detection server is configured to acquire whether the plurality of detection devices have a movable body in the designated area.
  • a detection result in the plurality of detection devices having a plurality of different types, each of the detection devices outputs a detection result; and determining, from the plurality of detection results, a plurality of input parameters are input into the detection model;
  • the detection model obtains an output parameter according to the determination weight corresponding to the type of the plurality of detection devices and the plurality of input parameters; and obtaining, according to the output parameter, a determination result of whether there is a movable body in the designated area;
  • the interaction end is configured to obtain a determination result of whether there is a movable body in the specified area, where the designated area has multiple detecting devices, and each of the plurality of detecting devices having multiple different types is detected
  • the device outputs a detection result, which is determined according to whether the plurality of detection devices have a plurality of detection results for whether there is a movable body in the designated area; if the determination result and the specified area have The actual result of the movable body is different, and an adjustment instruction is output, which is used to indicate that the determination result is erroneous.
  • the detection server is configured with a detection model for determining weights corresponding to different types of detection devices, and when multiple detection devices of different types are acquired, whether there are multiple detections of movable bodies in the specified area.
  • a plurality of input parameters may be determined from the plurality of detection results and input into the detection model. Since the detection result is from different types of detection devices, the detection device is determined according to the type of the detection device that outputs the detection result.
  • the detection model may obtain an output parameter according to the determined weight corresponding to the type of the detecting device and the plurality of input parameters determined according to the detection result of the detecting device, and the analysis may be performed by analyzing the output parameter Whether there is a determination result of the movable body in the area, it can be seen that the determination result of the specified area needs to refer to the detection results of a plurality of different types of detection devices, and even if the detection results of some types of detection devices are misjudged, it may be based on other Type of detection device output Correct detection results and arrive at a correct determination result, thereby minimizing detection of a single type of equipment due to technical limitations appear miscarriage of justice arising from the adverse effects on the user experience, and effectively improve the detection accuracy.
  • FIG. 1 is a schematic structural diagram of a detection environment according to an embodiment of the present invention
  • FIG. 2 is a flowchart of a method for detecting a method according to an embodiment of the present invention
  • FIG. 3 is a flowchart of a method for adjusting a weight according to an embodiment of the present invention
  • FIG. 4 is a flowchart of a method for detecting a method according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of a detection server according to an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of an interaction end according to an embodiment of the present disclosure.
  • FIG. 7 is a schematic structural diagram of a detection system according to an embodiment of the present invention.
  • detection techniques it has been possible to apply detection techniques to device control in buildings, for example by detecting sound to control the opening or closing of the lamps.
  • detection of a movable body may be included, and the movable body may be an object having a movable function, such as a person, an animal, a movable robot, or the like.
  • the use of various devices in buildings is often closely related to the movable body.
  • the lighting system, air conditioner and other devices in the office are generally turned on. When no one is present, devices such as lighting systems and air conditioners are generally turned off.
  • the detection accuracy of the movable body is of great significance to the automatic control of the building, which is one of the functions that the intelligent building hopes to achieve.
  • the detection of the movable body can detect whether there is a movable body in the designated area, that is, whether the movable body exists in the designated area, and whether or not the object is moving in the designated area, for example, one is detected.
  • a person reads a document through a computer in an office, it may be in a state of almost static for a long time. However, it is still necessary to provide lighting, air conditioning, and the like for the person. If the power device of the office is closed only according to the movement of no object, it is obvious that It is not suitable.
  • the designated area may be an area of a fixed range, for example, an office in an office building or the like.
  • the types of detection devices proposed in the embodiments of the present invention may be adjusted according to different scene requirements or calculation accuracy.
  • the infrared detection device and the acoustic wave detection device may be used as different types, or a single infrared sensor and an infrared sensor array may also be used. Different types.
  • the traditional movable body detection methods mainly include single detection methods such as infrared remote sensing detection and ultrasonic radar detection. Any one of the detection methods has its own technical limitations, so the accuracy of this single detection method is not high, often due to The technical limitations have been misjudged. For example, when an infrared remote sensing device is used to detect whether a movable body is present in a designated area, when the movable body is too close to the detecting device, there may be a case where the movable body is not detected, resulting in an error in the detection result.
  • an embodiment of the present invention provides a detection method, a related apparatus, and a system.
  • the detection server is configured with a detection model for determining weights corresponding to different types of detection devices, and when multiple detection devices of different types are acquired for a specified area.
  • the detecting model After having a plurality of detection results of the movable body, it may be determined from the plurality of detection results that a plurality of input parameters are input into the detection model, and since the detection result is from different types of detection devices, according to the output detection result
  • the type of the detecting device determines the determining weight corresponding to the detecting device type, and the detecting model can obtain the output parameter according to the determining weight corresponding to the detecting device type and the plurality of input parameters determined according to the detecting result of the detecting device, by using the
  • the analysis of the output parameter can obtain a determination result of whether there is a movable body in the specified area.
  • the determination result of the specified area needs to refer to the detection results of a plurality of different types of detection devices, even if some of the types of detection devices are Mistakes in the test results, it is also possible According to the correct detection results output by other types of detection equipment, the correct determination result is obtained, thereby minimizing the adverse effect of the single type detection equipment on the user experience due to the technical limitations, and effectively improving the The accuracy of the test.
  • the embodiment of the present invention is mainly implemented by a detection server.
  • the detection server 100 is configured with a detection model 200 having a determination weight corresponding to different types of detection devices. Taking three detecting devices, and the types of the three detecting devices are different, the three detecting devices may be disposed in a designated area 300, and the three detecting devices are respectively a laser positioning sensor 301 and an infrared human body occupancy detector. 302 and the noise collecting sensor 303, when detecting whether there is a movable body in the designated area 300, the three detecting devices can obtain corresponding three detection results, and the detection result obtained by the laser positioning sensor 301 is used as an input parameter. 1.
  • the detection result obtained by the infrared human occupancy detector 302 as the input parameter 2 and the detection result obtained by the noise collection sensor 303 are input as the input parameter 3 to the detection model 200, since the detection model 200 has different types of detection devices.
  • Corresponding determination weights for example, a determination weight 1 corresponding to the type of the laser positioning sensor, a determination weight 2 corresponding to the type of the infrared human occupancy detector, corresponding to the determination weight 3 of the type of the noise collection sensor, then the detection model 200 can be based on Determine the weight 1 as the weight corresponding to the input parameter 1, and determine the weight 2 as the input. 2 parameters corresponding weight, the weight is determined as the weight 3 calculated by the weight corresponding to the 3 input parameters an output parameter, and thus the output parameter is determined whether the determination result for the movable body in the designated region.
  • FIG. 2 is a flowchart of a method for detecting a method according to an embodiment of the present invention, which is applied to a detection server, and the method includes:
  • S201 Acquire a plurality of detection results of whether the plurality of detection devices have a movable body in the designated area, and each of the plurality of detection devices having the plurality of different types outputs one detection result.
  • detecting devices for detecting whether or not a movable body is available, for example, a human body occupancy sensor, a positioning sensor, a noise sensor, and the like.
  • Each type of testing device has its advantages and limitations. It can be understood as a limitation of the detection technology of a type of testing device, for example, for people who are stationary for a long time. Since the sound is not emitted for a long time, or the sound is small, the sound detection is likely to result in an error detection result of judging no one.
  • the technical limitations of different types of detection devices are generally different. Therefore, compared with the detection of a single type of detection device, it is detected by multiple types of detection devices whether or not the same designated area has a movable body.
  • the results of testing by various types of testing equipment will reduce the impact of the limitations of a single type of testing equipment to a certain extent, and improve the accuracy. Therefore, it is possible to detect whether there is a movable body in a specified area by using a plurality of different types of detecting devices, wherein the number of detecting devices can be determined according to actual needs, but the number of different types of detecting devices is at least two.
  • the number of types of detecting devices may be smaller than the number of detecting devices, for example, five detecting devices include three different types, one of which is a detecting device of type a, and one detecting device of type b, Three types of c detection equipment. Or, for a plurality of detecting devices, the number of types of detecting devices included may also be equal to the number of detecting devices, for example, five detecting devices include five different types, or the types of the five detecting devices are different. .
  • the specified area is detected by a plurality of different types of detecting devices, and the plurality of detecting devices may be installed in the designated area.
  • the plurality of detecting devices may be installed in the designated area.
  • three different types of testing devices detect the presence of someone in an office, and all three testing devices can be installed in the office.
  • one detecting device When detecting whether there is a movable body in a specified area, one detecting device can obtain a detection result, and there are two cases of the detection result, which may be that there is a movable body in the designated area, or there is no movable body in the designated area. .
  • Several detection devices can detect several corresponding detection areas, and these test results should be made in the same time period. For example, ten different types of testing equipment are installed in an office area, and if one of the offices is tested at the same time or sequentially in one time period, ten test results can be obtained.
  • the embodiment of the present invention does not limit the manner in which the detection server obtains multiple detection results, and may be connected to the control system of the plurality of detection devices by the detection server. Each time the detection is performed, the detection server acquires multiple detection results through the control system.
  • S202 Determine, from the plurality of detection results, that a plurality of input parameters are input into the detection model.
  • the detection is performed by the multiple detection devices. Since each detection device outputs a detection result, the detection result obtained by the detection server may be the same as the number of detection devices, but whether the specified area has The final result of the moving body is only one case, either with a movable body or without a movable body. After the detection server obtains multiple detection results, it is necessary to combine the multiple detection results for analysis and processing, thereby obtaining a final result.
  • the detection server is configured with a detection model, and the detection model can be used to analyze the detection result. Therefore, the detection server can determine the plurality of acquired detection results and input a plurality of input parameters into the detection model, so as to pass the detection model.
  • the multiple detection results are analyzed and processed.
  • the detection result obtained by the detection server may be the detection result in the specified area or the movable area in the specified area. If the detection result is directly input into the detection model, the detection model may not recognize the detection result. Therefore, the detection result can be converted into an input parameter input into the detection model.
  • the input parameter may be a parameter that the detection model can recognize, and is a parameter corresponding to the detection result, which can be converted by the detection result, and the input parameter can be represented by a numerical value, for example, having a movable body in a specified area.
  • This detection result can be identified by the parameter a, and the detection result for the absence of the movable body in the designated area can be identified by the parameter b. Which parameter is used is not limited.
  • “1” and “0” are input parameters, once During the detection, if the detection result of the detecting device A is that there is a movable body, and the detection result of the detecting device B and the detecting device C is that there is no movable body, the corresponding input parameter may be “1, 0, 0”.
  • the embodiment of the present invention provides a plurality of manners for determining a plurality of input parameters from a plurality of detection results according to a relationship between the number of the plurality of detection results and the number of the plurality of input parameters.
  • the number of the plurality of detection results is the same as the number of the plurality of input parameters, and both are N.
  • the N detection results may be input to the detection model as N input parameters.
  • the types of N detecting devices that obtain the N detection results can have at least two possible combinations:
  • One combination may be that the number of detecting device types is plural, but smaller than the number N of detecting devices, that is, the detecting devices of the same type among the N detecting devices, for example, for example, five detecting devices include one type a. Detection equipment, one type b detection equipment, three types c detection equipment. In this combination mode, when the determination result is calculated by the detection model, the input parameters determined by the detection results of the same type of detection device can use the same determination weight. In another combination, the number of device types can be detected to be the same as the number of detection devices, that is, the types of the N detection devices are different.
  • the plurality of detection results are specifically N, and the plurality of input parameters are specifically M, and M is less than N, that is, the number of detection results is more than the number of input parameters. In this case, it is necessary to reduce the number of detection results. Since the same type of detection equipment has the same technical defect, reducing the detection result of the same type of detection equipment has the least influence on the calculation of the determination result, so the same type of detection can be performed.
  • the test results of the device are aggregated to reduce the number of test results.
  • Y detection results corresponding to the detection devices of the type may be determined from the detection results of the X detection devices of one type, thereby determining M detection results from the N detection results.
  • X is greater than Y
  • X is less than N
  • Y is less than M.
  • Y can take 1.
  • the method for determining the Y detection results from the X detection results is not limited in the present invention.
  • Y detection results may be arbitrarily selected from the X detection results, or the movable detection bodies may be included in the X detection results. Proportion, determine Y test results with the same or similar ratio.
  • M detection results can be finally determined from the N detection results, so that the M detection results can be input into the detection model as M input parameters.
  • the detection model obtains an output parameter according to the determining weight corresponding to the type of the plurality of detecting devices and the plurality of input parameters.
  • the detection model has a determination weight corresponding to different types of detection devices, that is, each detection device type has a corresponding determination weight. Determining the weight may be used to indicate the proportion of the detection result determined by the detection device of one type in the detection result determined by the plurality of types of detection devices, and determining the weight may be a specific value, such as a decimal number, a percentage, etc., determining The greater the weight, the greater the proportion of the detection result of the detection device corresponding to the determined weight in the detection results of the plurality of different types of detection devices, that is, the more influence the detection result has on the final determination result. Big.
  • the determination weight in the detection model can be preset or dynamically adjusted.
  • determining the weight may be preset according to the number of detecting devices. For example, if three different types of detecting devices detect whether the specified area has a mobile device, the average value may be used to detect each type. The determination weight corresponding to the device is set to 1/3.
  • Determining the weight can also be calculated by analyzing the results of multiple historical detections. For example, there are three different types of detection devices, and the detection device A, the detection device B, and the detection device C detect whether the designated area has a mobile device, and each Perform a test to compare the test results of three different types of test equipment with the actual results, taking the movable body as an example.
  • the test result of the test device A is that no one, the test device B, and the test device C are the test results.
  • the actual result is no one, it means that the detection result of the detection device A is correct, the detection results of the detection device B and the detection device C are wrong, and the correct and error conditions of the detection results of each device are recorded, and so on. Multiple tests can be performed.
  • the number of times the detection result in the detection device A is correct can be counted, the number of times the detection result in the device B is detected correctly, and the number of times the detection result in the device C is correct is correct.
  • the sum of the correct number of detection results can get a total value, and the detection of each detection device
  • the determined weight corresponding to the result can be divided by the total number of times the detection device detects the correct result.
  • the number of times the detection result in the detection device A is correct is 80
  • the number of times the detection result in the detection device B is correct is 50
  • the determination weights of the three different types of detection devices can also be dynamically adjusted based on the increase in the number of samples calculated by the analysis.
  • the detection result of each detecting device is not used as a basis for dynamically adjusting the weight for each detection.
  • the detection result of each detecting device may be used as a basis for adjusting the determining weight when the determined result is inconsistent with the actual result. If the determined result is consistent with the actual result, it is proved that the current determined weight can be used to obtain the correct determination result, so that the existing determined weight can be maintained. constant.
  • the detecting server determines, from the obtained plurality of detection results, that a plurality of input parameters are input into the detection model, and the detection model may process the plurality of input parameters according to the determination weights corresponding to the plurality of detection device types respectively.
  • An output parameter the output parameter can be a specific value, which can be calculated by the detection model based on the input parameters and the determined weights.
  • a feasible way of calculating the output parameter may be to multiply a plurality of input parameters by their corresponding determination weights to obtain a plurality of product values, and then add the plurality of product values to obtain an output parameter, to three
  • the input parameters of the different types of detection device A, detection device B, and detection device C are 2/5, 2/5, and 1/5, respectively.
  • the detection result of device A is detected.
  • the determination result may be the final result obtained by one detection process, which may be understood as a final result obtained by the detection server comprehensively analyzing the plurality of detection results. There are two cases of the determination result, which may be that there is a movable body in the designated area, or there is no movable body in the specified area.
  • An output parameter can be obtained in S203.
  • it is necessary to determine whether there is a movable body in the designated area according to the output parameter. Therefore, it is necessary to analyze the obtained output parameter to obtain a certain determination result.
  • the specific determination manner is not limited in the present invention.
  • the value of the movable body in the specified area may be used as a basis for obtaining the determination result according to the input parameter. If the output parameter is close to the evaluation value, it may be considered according to the output parameter.
  • the result of the determination is that there is a movable body in the designated area, and vice versa, the determination result that there is no movable body in the specified area is obtained.
  • the embodiment of the present invention provides a specific manner of obtaining a determination result according to an output parameter.
  • the output parameter may be a specific value.
  • the corresponding data may also be set.
  • the first data may be used to indicate that the specified area has a movable body
  • the second data may be used to represent the specified area. There is no movable body in it, and it is determined whether there is a movable body in the designated area by judging how close the output parameter is to the first data and the second data.
  • the degree of proximity may be used to indicate the correlation of the output data with the first data and the second data
  • the first data may be used to identify data having a movable body
  • the second data may be used to identify data without a movable body when the output
  • the proximity of the data to the first data is higher than the second data, indicating that the output data has a higher correlation with the first data
  • there is a higher possibility that the designated area has a movable body that is, the determination result may be
  • the specified area has a movable body; when the proximity of the output data to the second data is higher than the first data, indicating that the output data has a higher correlation with the second data, then there is no movable body in the designated area.
  • the possibility is higher, that is, the determination result may be that there is no movable body in the specified area.
  • the specific values of the first data and the second data may be set according to the numerical range of the output parameter, wherein the numerical range of the output parameter may be obtained according to the input parameter.
  • the input parameter is represented by two values of "100” and "0”
  • "100” indicates that the detection result has a movable body
  • "0" indicates that the detection result has no movable body
  • the output parameter can be a value in the range of 0-100, so the first data can be set to "100" and the second data set to "0".
  • the manner of determining the degree of proximity is not limited.
  • a feasible manner may be to use the difference to calculate the proximity, and calculate the difference between the output parameter and the first data and the output parameter and the second data, respectively.
  • the first data is 100
  • the second data is 0,
  • the output data is 70
  • the difference between the output data and the first data is 30, and the difference between the output data and the second data is 70
  • the output data is If the difference between the data and the difference between the data and the second data is smaller, the output data is more closely related to the first data, that is, the output data is more correlated with the first data, and the first data may be used.
  • the detection precision is high, the difference between the output parameter and the first data needs to be very small to have a high degree of closeness, for example, The first data is 100, the second data is 0, the output data is 70, and the output data needs to be equal to the first data difference of at most 10 to determine that the output data is relatively closer to the first data.
  • the case of 70 will be judged to be relatively close to the second data, and since the second data may be data for identifying the result of "no movable body", there is a possibility that there is no movable body in the designated area. It is higher, so it can be concluded that there is no movable body in the specified area.
  • the detection server is configured with a detection model for determining weights corresponding to different types of detection devices, and when multiple detection devices of different types are acquired, whether there are multiple detections of movable bodies in the specified area.
  • a plurality of input parameters may be determined from the plurality of detection results and input into the detection model. Since the detection result is from different types of detection devices, the detection device is determined according to the type of the detection device that outputs the detection result.
  • the detection model may obtain an output parameter according to the determined weight corresponding to the type of the detecting device and the plurality of input parameters determined according to the detection result of the detecting device, and the analysis may be performed by analyzing the output parameter Whether there is a determination result of the movable body in the area, it can be seen that the determination result of the specified area needs to refer to the detection results of a plurality of different types of detection devices, and even if the detection results of some types of detection devices are misjudged, it may be based on other Type of detection device output Correct detection results and arrive at a correct determination result, thereby minimizing detection of a single type of equipment due to technical limitations appear miscarriage of justice arising from the adverse effects on the user experience, and effectively improve the detection accuracy.
  • the determination result does not coincide with the actual result
  • the result of the determination is that there is no movable body in the designated area
  • the actual result is that there is a movable body in the designated area.
  • the determination result is obtained, it can be compared with the actual result. If the determination result is different from the actual result, it indicates that the determination result has an error.
  • a feedback operation on the determination result may be added to the case where the determination result is erroneous.
  • the method further includes:
  • the determination result is different from the actual result of whether the movable area has the movable body in the specified area, it is determined whether the N detection results are the same as the actual result; if the detection result is the same as the actual result, Corresponding weights corresponding to the types of detection devices of the detection results that are the same as the actual results in the detection model are correspondingly increased.
  • the actual result here can be understood as the fact that there is a movable body in the designated area in the period in which the determination result is obtained.
  • the actual results can be verified by verification, or by means of machine verification. For example, if someone in an office is tested, it can be checked by arranging staff to check whether the office actually has someone. Or monitor the actual situation of the office through the monitoring system.
  • the actual results can also be derived from relevant operations in the designated area. For example, if someone in an office is tested and the result is that there is no one in the office, the office's lighting control system can automatically turn off the lighting, in the office. Actually, some people, such as the staff, turn off the lighting device and obviously interfere with the normal work of the staff. The worker can turn the lighting device back on by the control switch, so if it is detected that the closed lighting device is quickly reopened, It can be reflected that the actual result is that someone in the office can judge that the determination result is different from the actual result by the lighting device being turned off and the lighting device being reopened.
  • the determination result is obtained by comprehensively analyzing the plurality of detection results by the detection server.
  • the determination result is different from the actual result, the detection result of the error in the plurality of detection results is indicated, and the errors are The result of the test has a great influence on the determination result, which leads to the error of the final determination result.
  • Determining the weight is an important factor reflecting the influence of the detection result on the determination result.
  • the detection device type that obtains the error detection result is higher than the determination weight corresponding to the detection device type that obtains the correct detection result, the detection result is obtained.
  • the type of the detection device that correctly detects the result is lower than the determination weight corresponding to the type of the detection device that obtains the error detection result, the erroneous detection result has a greater influence on the determination result, thereby causing the determination result to be inconsistent with the actual result.
  • there are three different types of detection devices, the detection device A, the detection device B, and the detection device C, and the corresponding determination weights are all 1/3. In one test, the detection result of the device A is detected by comparing with the actual result.
  • the detection results of the detection device B and the detection device C are incorrect. It can be known that the determination weight corresponding to the detection device type of the detection result in the detection result is 1/3, and the determination weight corresponding to the detection device type of the error detection result is 2/. 3. In contrast, since the determination weight of the detection device A is low due to the correct detection result, the final determination result is consistent with the detection results of the detection device B and the detection device C, that is, the determination result does not match the actual result.
  • the detection device type that obtains the correct detection result has a lower determination weight than the detection device type that obtains the error detection result, and is a main factor leading to an error in the determination result. If the determination weight corresponding to the type of the detection device has not changed, when a similar detection scenario occurs next time, the determination result still has an error that does not match the actual result. Therefore, when the determination result is different from the actual result, the occurrence of the above situation can be avoided by increasing the determination weight corresponding to the type of the detection device that obtains the same detection result as the actual result.
  • the detection result obtained by the detection device of the type can play a greater role in the determination result, thereby making the determination result and the detection.
  • the result is equally likely to be the same, that is, the probability that the result is consistent with the actual result is higher.
  • the determination weight corresponding to one type of detection device can be obtained by the proportion of the detection result of the detection device in the detection results of the plurality of different types of detection devices, the plurality of different types
  • the determination weights corresponding to the detection devices are generally a fixed value. Therefore, when the determination weight corresponding to the detection result in the detection model is increased, the detection result corresponding to the actual result in the corresponding detection model corresponds. The determination weight will be reduced.
  • the difference is used to calculate the proximity.
  • the determination weights corresponding to different types of detection device A, detection device B, and detection device C are both 1/3, and all determination weights are added to 1.
  • the detection result of detection device A is the same as the actual result, and detection device B and The detection result of the detection device C is different from the actual result.
  • the determination weight corresponding to the detection device A having the same detection result as the actual result is 1/3, and the detection device B and the detection device C having different detection results from the actual result are obtained.
  • the sum of the determined weights is 2/3, and the final result is the same as the detection result of the detecting device B and the detecting device C, that is, the final result does not match the actual result, in order to make the final determination when the above detection result occurs again.
  • the result can be consistent with the actual result, instead of regaining the determination result that is inconsistent with the actual result. Therefore, it is necessary to increase the determination weight corresponding to the detection device A, for example, to adjust it to 3/5, and the detection device B and the detection device C correspond to each other. The determined weight will be reduced to 2/5. Therefore, when the above detection result occurs again, the final determination result is consistent with the actual result.
  • the determination weight corresponding to the detection device type that obtains the same detection result as the actual result in the detection model can be increased, thereby optimizing the detection model, thereby further improving the detection accuracy. Sex.
  • the specific manner of determining the weight is introduced.
  • the actual detection result of the multiple different types of detection devices may be represented by establishing a data sequence.
  • each data sequence includes the plurality of data bits, each data bit corresponding to a type of detection device.
  • the parameters recorded in the data sequence are used as the basis for determining the weights in the dynamic adjustment detection model.
  • a data sequence including M data bits may be arranged in the order of the detecting device.
  • M different types of detecting devices are L1-LM in sequence, corresponding to A data sequence includes M data bits, the first data bit may correspond to the first detecting device L1, the second data bit may correspond to the second detecting device L2, and so on, the Mth data bit may correspond At the Mth detecting device LM.
  • a data sequence including M data bits may also be randomly arranged.
  • three different types of detection devices are sequentially L1-L3, the first data bit may correspond to the first detection device L1, and the second data bit corresponds to The third detecting device L3, the third data bit corresponds to the second detecting device L2.
  • the determination result is the same as the actual result, it can be considered that the currently used determination weight configuration is relatively reasonable, and the correct determination result can be obtained.
  • the determination result is different from the actual result, the currently used determination weight configuration can be considered. The possibility may be unreasonable, resulting in a erroneous test result having a greater impact on the whole, and finally a erroneous determination result. Therefore, when a erroneous determination result occurs, a new data sequence, that is, a target sequence, is created, and the detection results obtained by the respective detection devices are correspondingly filled into the target sequence, and the detection result is implemented according to the target sequence. Determine the adjustment of the weight, as shown in Figure 3, the specific operation is as follows:
  • S301 Establish a target sequence according to the determination result of determining whether the plurality of detection results are the same as the actual result, where the target sequence is a data sequence, and in the target sequence, outputting the same result as the actual result
  • the detection data of the detection device is filled with a valid factor in the data bit, and the other data bits are filled with an invalid factor.
  • the correctness of the detection result of the detecting device can be indicated by filling the effective factor or the invalid factor in each data bit.
  • the effective factor can be used to indicate that the detection result of the detecting device is the same as the actual result, and the invalidity factor can be used to indicate that the detection result of the detecting device is different from the actual result, and the effective factor and the invalid factor can be in various forms, which can be numbers, letters or The text of the embodiment of the present invention is not limited in its form. For example, if the number represents the effective factor and the letter represents the invalid factor, then the factor filled in the data bit is the effective factor in the form of a number, and the factor filled in the data bit is the invalid form in the form of a letter.
  • a number less than 100 represents an effective factor
  • a number greater than or equal to 100 represents an invalid factor
  • a factor filled in the data bit is a number less than 100, which is an effective factor
  • a factor filled in the data bit is greater than or equal to The number of 100 is the invalid factor.
  • the detection model may have a plurality of data sequences, and the data in the plurality of data sequences may be pre-set test data, for example, by presetting various types of scenes that are difficult to identify, so as to achieve a situation in which the determination result is inconsistent with the actual result. Therefore, the detection result obtained by each detecting device when the determination result is inconsistent with the actual result is filled in the data sequence as test data.
  • the data in the plurality of data sequences may also be randomly filled data, and used as a basis for determining the weights after the loop iteration calculation.
  • the data in the plurality of data sequences may also be data corresponding to the detection results obtained by the respective detecting devices when the determining result does not match the actual result in the normal detecting process.
  • detection device A different types detect whether a person in the designated area is detected, and the detection result of the detection device A is that the detection result of the person, the detection device B, and the detection device C are unmanned, according to this.
  • the result of the three test results is that there is no one in the designated area, and the actual result of the designated area is someone, and the visible result is inconsistent with the actual result.
  • the detection result of the detection device A is the same as the actual result, and the detection result of the detection device B and the detection device C is different from the actual result, and a data sequence “1, 0, 0” can be established according to the detection results of the three detection devices.
  • the target sequence can reflect the correctness and error of multiple detection results compared with the actual results when the determination result is different from the actual result. Since the adjustment of the determination weight is related to the number of effective factors, the establishment of the target sequence can be used to increase the number of effective factors of the data bits corresponding to the detection device having the same detection result as the actual result.
  • different types of detection device A, detection device B, and detection device C detect whether a person in a specified area is detected, and the detection result of the detection device A is unmanned, and the detection results of the detection device B and the detection device C are all people, and the detection device
  • the determination weight corresponding to A is higher than the determination weight corresponding to the detection device B and the detection device C, and the determined result is no one, but the actual result is someone, and the final determination result is inconsistent with the actual result, due to The detection result of the detection device A is different from the actual result.
  • the detection result of the detection device B and the detection device C is the same as the actual result, and the established target sequence is “0, 1, 1”, in order to improve the detection device B to a certain extent.
  • the type and the type of the detection device C correspond to the determined weight.
  • different types of detecting device A, detecting device B, and detecting device C detect whether a person in the designated area is detected, and the detection result of the detecting device A is that the detection result of the person, the detecting device B, and the detecting device C are unmanned, and the detecting is performed.
  • the determining weight corresponding to the device A is higher than the determining weight corresponding to the detecting device B and the detecting device C, and the determined result is a person, but the actual result is no one, and the final determined result does not match the actual result.
  • the target sequence established is “0, 1, 1”, in order to improve the detection device B to a certain extent.
  • the type and the type of detection device C correspond to the determined weights.
  • a feasible method for determining the determining weight corresponding to the first device may be: the total number of valid factors filled in the data bits corresponding to the first device, and all data sequences in the detection model. The total number of valid factors filled is divided and divided, as shown in Table 1:
  • Table 1 "1" represents the effective factor, "0" represents the invalid factor, and the first row in Table 1 represents different types of detection devices, which are the first device, the second device, and the third device, respectively, the second row, the first
  • the three rows and the fourth behavior detection model have three data sequences, the fifth row is the established target sequence, each column data is data corresponding to the detection device, and the first column data is data corresponding to the first device.
  • the determination result obtained by combining the three detection results is unmanned.
  • the first device that is equivalent to the detection result that is consistent with the actual result has less influence on the determination result in the three detection devices, and needs to increase the determination weight corresponding to the first device, thereby inputting
  • the target sequence, the data sequence shown in line 5, can be "1, 0, 0".
  • the target sequence that is, the data sequence shown in the 5th line, can be "1, 0, 0".
  • the first device (the detection device with the same detection result as the actual result) is taken as an example to introduce the adjustment of the weight determination.
  • the data bit of the detecting device is filled with an invalid factor, and the total number of valid factors filled in the data bits is not changed, but due to the same detection result as the actual result.
  • the data bits of the detecting device are filled with a valid factor, so that the total number of valid factors filled in all the data sequences changes, so the determining weight corresponding to the detecting device type of the detection result different from the actual result also changes correspondingly.
  • the method for calculating the weight is similar to the method for determining the weight corresponding to the first device, and details are not described herein again.
  • Table 2 shows:
  • Testing equipment A Testing equipment B Testing equipment C 1 0 0 0 1 0 0 0 1 0 1 0 1 1 1
  • the first row in the table represents different types of detection devices, namely detection device A, detection device B and detection device C, second row, third
  • the row and the fourth behavior detection model have three data sequences, the fifth row is the established target sequence, each column data is data corresponding to the detection device, the first column data is the data corresponding to the detection device A, and the second column data It is the data corresponding to the detection device B, and the third column data is the data corresponding to the detection device C.
  • the detection result of the detecting device A is a person and the detection result of the detecting device B and the detecting device C is unmanned for a designated area, if the result of the synthesis of the three detection results is a person, but When the actual result is unmanned, the detection device B and the detection device C which are equivalent to the detection result that is consistent with the actual result have less influence on the determination result in the three detection devices, and the detection device B and the detection device C need to be increased. Correspondingly determining the weight, so that the target sequence entered, that is, the data sequence shown in the fifth row can be "0, 1, 1".
  • the detection result of the detecting device A is unmanned, and the detection result of the detecting device B and the detecting device C is a person, if the result of the synthesis of the three detection results is unmanned
  • the detecting device B and the detecting device C which are equivalent to the detection result which is consistent with the actual result have less influence on the determination result in the three detecting devices, and it is necessary to increase the detecting device B and the detecting device.
  • the weight corresponding to C is determined so that the target sequence entered, that is, the data sequence shown in the fifth row can be "0, 1, 1".
  • the total number of valid factors filled in all data sequences of the detection model is five, and in all data sequences, the total number of valid factors filled in the data bits corresponding to the detection device A is one, then detection The determining weight corresponding to the device A is 1/5, and the total number of valid factors filled in the data bits corresponding to the detecting device B is two, and the determining weight corresponding to the detecting device B is increased from 1/3 to 2/5.
  • the total number of valid factors filled in the data bits corresponding to the detecting device C is two, and the determining weight corresponding to the detecting device C is increased from 1/3 to 2/5.
  • the adjustment of the weight is mainly determined according to the number of the effective factor and the invalid factor, that is, the form of the effective factor and the invalid factor have no influence on the adjustment of the determined weight, and therefore, even for the same
  • the number of effective factors of the detection device that increases the detection result identical to the actual result is used to increase the detection device type corresponding to the detection result of the actual result.
  • the weight is determined such that the detection result of the detecting device can have a higher degree of influence on the determination result, thereby further improving the accuracy of the determination result.
  • the interactive terminal described herein may be a smart device having an interactive function, such as a mobile phone, a PAD, a personal computer, or the like.
  • the interaction end has a function of establishing a data connection with the detection server, so that the user can know the determination result for the specified area through the interaction end, and can indicate the determination result error to the detection server through the interaction end when the actual result is different from the actual result. .
  • FIG. 4 is a flowchart of a method for detecting a method according to an embodiment of the present invention, which is applied to an interaction end, where the method includes:
  • S401 Acquire a determination result of whether there is a movable body in the designated area, where the designated area has a plurality of detecting devices, and each of the plurality of detecting devices having a plurality of different types outputs one detecting result The determination result is determined according to whether the plurality of detecting devices have a plurality of detection results for whether there is a movable body in the designated area.
  • the determination result may be sent by the detection server, or may be sent by other devices, and the present invention is not limited thereto.
  • the interaction terminal can clarify whether the determination result indicates whether there is a movable body in the designated area.
  • the actual result may be obtained by user observation of the interactive end, or may be determined by a related operation generated in the designated area. If the determined result is a non-movable body, the related operation may be, for example, suddenly after the power is turned off. The specified area is opened.
  • the interaction end may generate an adjustment instruction and output, for example, send the adjustment instruction to the detection server to notify the detection server that the determination result made for the designated area is wrong. . Therefore, the detection server can adjust the determination weights corresponding to different detection device types according to the relationship between the detection result and the actual result, so as to improve the determination weight of the type of the detection device that is the same as the actual result. For details, refer to the related description in the embodiment corresponding to FIG. 2 and FIG. 3, and details are not described herein again.
  • the interaction end may control the specified area according to the determination result. Controllable device.
  • the controllable device described herein may refer to a device that can be remotely controlled, and may mainly include a device that provides a service function in the designated area, for example, the controllable device includes any one or more of a lighting device, an air conditioner, a display device, and a projection device. The combination.
  • the interaction segment can close the controllable device in the designated area, such as lighting, air conditioning, and the like.
  • the interaction segment can control the controllable device in the designated area to be in a service state, for example, the lighting, air conditioner, etc. to be turned off. turn on.
  • the verification of the determination result by the interaction end can effectively reduce the adverse effect of the erroneous determination result on the designated area, and provide feedback for the detection server to adjust the calculation and determination result through feedback of the verification result, so as to improve The accuracy of the results was determined to contribute.
  • FIG. 5 is a detection server according to an embodiment of the present invention.
  • the detection server is configured with a detection model, and the detection model has a determination weight corresponding to different types of detection devices, and the detection server includes an acquisition unit 501 and Determination unit 502:
  • the acquiring unit 501 is configured to acquire, by the plurality of detecting devices, a plurality of detection results for whether there is a movable body in the designated area, and each of the plurality of detecting devices that output the plurality of detection devices outputs one detection. result;
  • the determining unit 502 is configured to determine, from the plurality of detection results, that a plurality of input parameters are input into the detection model;
  • the determining unit 502 is further configured to use the detection model to obtain an output parameter according to the determining weight corresponding to the type of the plurality of detecting devices and the plurality of input parameters;
  • the determining unit 502 is further configured to obtain, according to the output parameter, a determination result of whether there is a movable body in the designated area.
  • the detection server further includes a determining unit and an adding unit:
  • the determining unit is configured to determine whether the plurality of detection results are the same as the actual result. ;
  • the adding unit is triggered, and the adding unit is configured to correspondingly increase the determining weight corresponding to the detecting device type of the detection result that is the same as the actual result in the detection model.
  • the detection model has a plurality of data sequences, each of the data sequences includes M data bits, each data bit corresponding to one type of detecting device, and the adding unit is further configured to: according to the determining Whether the detection result is the same as the actual result in the detection result, the target sequence is established, and the target sequence is a data sequence, and the target device outputs the corresponding detection device corresponding to the actual result.
  • the data bits are filled with valid factors, and other data bits are filled with invalid factors;
  • Adjusting the total number of valid factors filled in the data bits of all the data sequences of the detection model and the data bits corresponding to the detection devices outputting the same detection result as the actual result in all data sequences Determining a weight corresponding to the detection device type of the detection result that is the same as the actual result; wherein, for the first device, the total number of valid factors filled in all data sequences according to the detection model, and the first in all data sequences Determining, by the total number of valid factors filled in the data bits corresponding to the device, the determining weight corresponding to the first device, where the first device is detecting any one of the detecting devices that output the detection result with the same actual result device.
  • the determining unit includes a determining subunit and determining the subunit:
  • the determining subunit is configured to determine a degree of proximity between the output data of the detection model and the first data and the second data, where the first data is used to identify data having a movable body, and the second The data is data for identifying that there is no movable body;
  • the determining subunit is triggered, and the determining subunit is configured to obtain a determination result that the movable area is in the designated area;
  • the determining subunit is triggered, and the determining subunit is configured to obtain a determination result that there is no movable body in the designated area.
  • the type of the detecting device includes at least a human body occupancy sensor, a positioning sensor, and a noise sensor.
  • the multiple detection results are specifically N
  • the multiple input parameters are specifically N
  • the determining unit is further configured to input the N detection results as the N input parameters to the detection.
  • the model In the model.
  • the types of the N detecting devices are different.
  • the multiple detection results are specifically N
  • the multiple input parameters are specifically M
  • M is smaller than N
  • the N detecting devices have M different types
  • the determining unit is further used to The detection results corresponding to the detection devices of this type are determined by the detection results of the X detection devices of one type, so that M detection results are determined from the N detection results, and X is greater than Y.
  • X is less than N
  • Y is less than M;
  • the M detection results are input to the detection model as M input parameters.
  • a detection model for determining weights corresponding to different types of detection devices is configured, and after obtaining a plurality of detection methods of the different types of detection devices for the specified area, the A plurality of input parameters are determined to be input into the detection model. Since the detection result is from different types of detection devices, the determination weight corresponding to the detection device type is determined according to the type of the detection device that outputs the detection result.
  • the detection model can obtain an output parameter according to the determination weight corresponding to the detection device type and the plurality of input parameters determined according to the detection result of the detection device, and the analysis of the output parameter can be used to determine whether there is a movable in the specified region.
  • the determination result of the specified area needs to refer to the detection results of a plurality of different types of detection devices, and even if the detection results of some types of detection devices are misjudged, it may be output according to other types of detection devices.
  • Correct test result The right to determine the results, thereby minimizing detection of a single type of equipment due to the negative impact on the user experience technical limitations arising from miscarriage of justice, and effectively improve the detection accuracy.
  • FIG. 6 is an interaction end according to an embodiment of the present disclosure, where the interaction end includes an obtaining unit 601 and an output unit 602:
  • the obtaining unit 601 is configured to acquire a determination result of whether there is a movable body in the designated area, where the specified area has a plurality of detecting devices, each of the plurality of detecting devices having a plurality of different types
  • the detecting device outputs a detection result, which is determined according to a plurality of detection results of the plurality of detecting devices for whether there is a movable body in the designated area;
  • the output unit 602 is triggered, and the output unit 602 is configured to output an adjustment instruction, where the adjustment instruction is used to indicate the determination. The result is wrong.
  • control unit also includes a control unit:
  • the control unit is triggered, and the control unit is configured to control the controllable device in the designated area according to the determination result.
  • controllable device comprises a combination of any one or more of a lighting device, an air conditioner, a display device, and a projection device.
  • the verification of the determination result by the interaction end can effectively reduce the adverse effect of the erroneous determination result on the designated area, and provide feedback for the detection server to adjust the calculation and determination result through feedback of the verification result, so as to improve The accuracy of the results was determined to contribute.
  • FIG. 7 is a detection system according to an embodiment of the present invention.
  • the detection system includes a detection server 701 and an interaction end 702:
  • the detection server 701 is configured with a detection model, and the detection model has a determination weight corresponding to different types of detection devices, and the detection server 701 is configured to acquire whether the plurality of detection devices have a movable body in the designated area. a plurality of detection results, each of the plurality of detection devices having a plurality of different types outputting a detection result; determining, from the plurality of detection results, inputting a plurality of input parameters to the detection model
  • the detection model obtains an output parameter according to the determination weight corresponding to the type of the plurality of detection devices and the plurality of input parameters; and determining, according to the output parameter, whether there is a movable body in the designated area result.
  • the interaction end 702 is configured to acquire a determination result of whether there is a movable body in the designated area, where the specified area has multiple detection devices, and each of the plurality of detection devices having multiple different types
  • the detecting device outputs a detection result, which is determined according to whether the plurality of detecting devices have a plurality of detection results for whether there is a movable body in the designated area; if the determination result is in the specified area
  • the actual result with the movable body is different, and an adjustment command is output, which is used to indicate that the determination result is erroneous.

Landscapes

  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Geophysics (AREA)
  • Air Conditioning Control Device (AREA)
  • Selective Calling Equipment (AREA)
  • Testing Or Calibration Of Command Recording Devices (AREA)

Abstract

本发明实施例公开了一种检测方法、相关装置和系统,检测服务器中配置有针对不同类型检测设备所对应的确定权重的检测模型,当获取到多个检测结果后,可以从该多个检测结果中确定出多个输入参数输入到所述检测模型中,检测模型根据检测设备类型所对应的确定权重以及根据检测设备的检测结果所确定的多个输入参数可以得到输出参数,通过对所述输出参数的分析可以得到针对所述指定区域中是否具有可移动体的确定结果,从而尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高了检测的准确性。

Description

一种检测方法、相关装置和系统
本申请要求2016年12月29日递交的申请号为201611247432.8、发明名称为“一种检测方法、相关装置和系统”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及数据处理领域,特别是涉及一种检测方法、相关装置和系统。
背景技术
随着检测技术的发展,已经可以将检测技术应用到对建筑物中器件控制中,例如常见的通过检测声音来控制打开的声控灯。除了这种简单的应用外,还可以包括对可移动体的检测,可移动体可以包括人、动物、可移动机器人等。
对可移动体的检测结果可以应用于建筑物中各类器件的控制上,例如当检测到一间会议室中没有可移动体时,系统可以将这间会议室的空调照明等器件关闭以便节能。可见,对可移动体的检测准确度对建筑物的自动化控制意义重大。
发明内容
传统的可移动体检测方式主要包括红外遥感检测、超声雷达检测等单一检测方式,任何一种检测方式都有自身的局限性,故导致这种单一检测方式的准确性不高,经常会出现误判的情况,例如将原本有人的房间判断为无人,导致系统在实际有人的情况下将该房间照明空调关闭,对该房间中的人产生了不好的用户体验。
为了解决上述技术问题,本发明提供了一种检测方法、相关装置和系统,可以尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高检测的准确性。
本发明实施例公开了如下技术方案:
第一方面,本发明提供了一种检测方法,应用于检测服务器,所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述方法包括:
获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;
从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;
所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;
根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
可选的,在所述根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果之后,还包括:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,判断所述多个检测结果中是否有与所述实际结果相同;
若具有与所述实际结果相同的检测结果,相应增加所述检测模型中所述与实际结果相同的检测结果的检测设备类型所对应的确定权重。
可选的,所述检测模型中具有多个数据序列,每一个数据序列包括M个数据位,每一个数据位对应一个类型的检测设备,所述相应增加所述检测模型中所述与实际结果相同的检测结果的检测设备类型所对应的确定权重,包括:
根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出所述与实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子;
根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出所述与实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的总数,调整所述与实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出所述与实际结果相同的检测结果的检测设备中的任意一个检测设备。
可选的,所述根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果,包括:
判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度,所述第一数据为用于标识具有可移动体的数据,所述第二数据为用于标识没有可移动体的数据;
若所述输出数据相对于第二数据更接近第一数据,得出所述指定区域中具有可移动体的确定结果;
若所述输出数据相对于第一数据更接近第二数据,得出所述指定区域中没有可移动 体的确定结果。
可选的,所述检测设备的类型至少包括人体占用传感器、定位传感器和噪声传感器。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为N个,所述从所述多个检测结果中确定出多个输入参数输入到所述检测模型中,包括:
将所述N个检测结果作为N个输入参数输入到所述检测模型中。
可选的,所述N个检测设备的类型各不相同。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M小于N,所述N个检测设备具有M个不同类型,所述从所述多个检测结果确定中出多个输入参数输入到所述检测模型中,包括:
从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M;
将所述M个检测结果作为M个输入参数输入到所述检测模型中。
第二方面,本发明提供了一种检测方法,应用于交互端,所述方法包括:
获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
可选的,还包括:若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,根据所述确定结果控制所述指定区域中的可控设备。
可选的,所述可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
第三方面,本发明提供了一种检测服务器,所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器包括获取单元和确定单元:
所述获取单元,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;
所述确定单元,用于从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;
所述确定单元还用于利用所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;
所述确定单元还用于根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
可选的,所述检测服务器还包括判断单元和增加单元:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述判断单元,所述判断单元,用于判断所述多个检测结果中是否有与所述实际结果相同;
若具有与所述实际结果相同的检测结果,触发所述增加单元,所述增加单元用于相应增加所述检测模型中所述与实际结果相同的检测结果的检测设备类型所对应的确定权重。
可选的,所述检测模型中具有多个数据序列,每一个数据序列包括M个数据位,每一个数据位对应一个类型的检测设备,所述增加单元还用于根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出所述与实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子;
根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出所述与实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的总数,调整所述与实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出所述与实际结果相同的检测结果的检测设备中的任意一个检测设备。
可选的,所述确定单元包括判断子单元和确定子单元:
所述判断子单元,用于判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度,所述第一数据为用于标识具有可移动体的数据,所述第二数据为用于标识没有可移动体的数据;
若所述输出数据相对于第二数据更接近第一数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中具有可移动体的确定结果;
若所述输出数据相对于第一数据更接近第二数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中没有可移动体的确定结果。
可选的,所述检测设备的类型至少包括人体占用传感器、定位传感器和噪声传感器。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为N个,所述确定单元还用于将所述N个检测结果作为N个输入参数输入到所述检测模型中。
可选的,所述N个检测设备的类型各不相同。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M小于N,所述N个检测设备具有M个不同类型,所述确定单元还用于从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M;
将所述M个检测结果作为M个输入参数输入到所述检测模型中。
第四方面,本发明提供了一种交互端,所述交互端包括获取单元和输出单元:
所述获取单元,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述输出单元,所述输出单元,用于输出调整指令,所述调整指令用于指示所述确定结果是错误的。
可选的,还包括控制单元:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,触发所述控制单元,所述控制单元,用于根据所述确定结果控制所述指定区域中的可控设备。
可选的,所述可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
第五方面,本发明提供了一种检测系统,所述检测系统包括检测服务器和交互端:
所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数 得到输出参数;根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果;
所述交互端,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
由上述技术方案可以看出,检测服务器中配置有针对不同类型检测设备所对应的确定权重的检测模型,当获取到多个不同类型的检测设备针对指定区域中是否具有可移动体的多个检测结果后,可以从该多个检测结果中确定出多个输入参数输入到所述检测模型中,由于检测结果来自不同类型的检测设备,故根据输出检测结果的检测设备的类型确定出该检测设备类型对应的确定权重,检测模型根据检测设备类型所对应的确定权重以及根据检测设备的检测结果所确定的多个输入参数可以得到输出参数,通过对所述输出参数的分析可以得到针对所述指定区域中是否具有可移动体的确定结果,可见,对指定区域的确定结果需要参考多个不同类型检测设备的检测结果,即使其中某些类型的检测设备的检测结果出现误判,也有可能根据其他类型的检测设备所输出的正确检测结果而得出正确的确定结果,从而尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高了检测的准确性。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。
图1为本发明实施例提供的一种检测环境的结构示意图;
图2为本发明实施例提供的一种检测方法的方法流程图;
图3为本发明实施例提供的一种调整确定权重的方法的流程图;
图4为本发明实施例提供的一种检测方法的方法流程图;
图5为本发明实施例提供的一种检测服务器的结构示意图;
图6为本发明实施例提供的一种交互端的结构示意图;
图7为本发明实施例提供的一种检测系统的结构示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
随着检测技术的发展,已经可以将检测技术应用到对建筑物中器件控制中,例如,通过检测声音来控制灯的打开或关闭。除了这种简单的应用外,还可以包括对可移动体的检测,所述可移动体可以是具有可移动功能的物体,例如,人、动物、可移动机器人等。在日常生活中,建筑物中各类器件的使用往往与可移动体有密切的关联,以人为例,当办公室中有人存在时,办公室中的照明系统、空调等器件一般处于开启状态,当办公室无人存在时,照明系统、空调等器件一般处于关闭状态。
因此,可以通过对可移动体是否存在进行检测,依据检测结果实现对建筑物中各类器件的自动化控制,例如,当检测到办公室有人存在时,可以通过智能系统控制办公室灯的打开,提高了用户的体验感,当检测到办公室无人存在时,则将灯关闭,可以起到节能作用。可见,对可移动体的检测准确度对建筑物的自动化控制意义重大,这也是智能建筑所希望实现的功能之一。
需要说明的是,对可移动体的检测可以在指定区域中检测是否具有可移动体,即检测在该指定区域中可移动体是否存在,并非检测在指定区域中是否有物体在移动,例如一个人在办公室内通过电脑阅读文档时,可能长时间处于几乎静止的状态,但是,依然需要为这个人提供照明、空调等服务,若仅根据没有物体移动而关闭了该办公室的用电设备,显然是不合适的。其中,指定区域可以是固定范围的一个区域,例如,写字楼中的一间办公室等。本发明实施例所提出的检测设备的类型可以根据不同的场景需求或者计算精度进行调整,例如,红外检测设备和声波检测设备可以作为不同的类型,或者,单红外传感器和红外传感器阵列也可以作为不同的类型。
传统的可移动体检测方式主要包括红外遥感检测、超声雷达检测等单一检测方式,任何一种检测方式都有自身的技术局限性,故导致这种单一检测方式的准确性不高,经常会由于技术局限性出现误判的情况。例如,采用红外遥感设备检测指定区域中是否具有可移动体时,当可移动体与该检测设备距离太近时,可能会存在检测不到可移动体的 情况,从而导致检测结果错误。
为此,本发明实施例提供一种检测方法、相关装置和系统,检测服务器中配置有针对不同类型检测设备所对应的确定权重的检测模型,当获取到多个不同类型的检测设备针对指定区域中是否具有可移动体的多个检测结果后,可以从该多个检测结果中确定出多个输入参数输入到所述检测模型中,由于检测结果来自不同类型的检测设备,故根据输出检测结果的检测设备的类型确定出该检测设备类型对应的确定权重,检测模型根据检测设备类型所对应的确定权重以及根据检测设备的检测结果所确定的多个输入参数可以得到输出参数,通过对所述输出参数的分析可以得到针对所述指定区域中是否具有可移动体的确定结果,可见,对指定区域的确定结果需要参考多个不同类型检测设备的检测结果,即使其中某些类型的检测设备的检测结果出现误判,也有可能根据其他类型的检测设备所输出的正确检测结果而得出正确的确定结果,从而尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高了检测的准确性。
本发明实施例主要通过检测服务器来实现,可以参见图1所示,检测服务器100中配置有检测模型200,检测模型200中具有针对不同类型检测设备所对应的确定权重。以三个检测设备,且这三个检测设备的类型各不相同为例,这三个检测设备可以设置于一个指定区域300中,三个检测设备分别为激光定位传感器301、红外人体占用检测器302和噪声收集传感器303,在对指定区域300中是否具有可移动体进行一次检测时,三个检测设备可以得出对应的三个检测结果,将激光定位传感器301得出的检测结果作为输入参数1、红外人体占用检测器302得出的检测结果作为输入参数2和噪声收集传感器303得出的检测结果作为输入参数3输入到检测模型200,由于检测模型200中具有与不同类型的检测设备所对应的确定权重,例如对应激光定位传感器这一类型的确定权重1,对应红外人体占用检测器这一类型的确定权重2,对应噪声收集传感器这一类型的确定权重3,那么检测模型200可以根据确定权重1作为输入参数1对应的权重,确定权重2作为输入参数2对应的权重,确定权重3作为输入参数3对应的权重计算出输出参数,并以此输出参数确定出针对所述指定区域中是否具有可移动体的确定结果。
接下来详细介绍本发明实施例所提供的检测方法。图2为本发明实施例提供的一种检测方法的方法流程图,应用于检测服务器,该方法包括:
S201:获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具 有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果。
对是否具有可移动体进行检测的检测设备,其类型可以有多种,例如,人体占用传感器、定位传感器和噪声传感器等。每种类型的检测设备都有其检测技术对应的优点,也有其的局限性,这里所述的局限性可以理解为一个类型的检测设备在检测技术上的局限性,例如针对长时间静止的人,由于长时间不发出声音,或者声音很小,声音检测容易出现判断无人的错误检测结果。
不同类型的检测设备的技术局限性一般来说有所区别,故相比于单一类型的检测设备的检测,通过多种类型的检测设备对同一个指定区域是否具有可移动体进行检测,综合多种类型的检测设备进行检测得出的结果会一定程度上减小单一类型检测设备的局限性所带来的影响,提高了准确性。故此,可以通过多个不同类型的检测设备对指定区域中是否具有可移动体进行检测,其中,检测设备的个数可以根据实际需求确定,不过不同类型的检测设备的个数至少为两个。对于多个检测设备,所包括检测设备的类型数量可以小于检测设备的个数,例如五个检测设备中包括三个不同的类型,其中有一个类型a的检测设备,一个类型b的检测设备,三个类型c的检测设备。或者,对于多个检测设备,所包括检测设备的类型数量也可以等于检测设备的个数,例如五个检测设备中包括五个不同的类型,或者说,这五个检测设备的类型各不相同。
在本发明实施例中通过多个不同类型的检测设备对指定区域检测,则可以将该多个检测设备都安装在该指定区域中。例如,三个不同类型的检测设备对一间办公室是否有人的情况进行检测,可以将这三个检测设备都安装在该办公室中。例如,也可以是将具有两种不同类型的三个检测设备设置在一个办公室中继续是否有人的检测。
在对指定区域中是否具有可移动体进行一次检测时,一个检测设备可以得到一个检测结果,检测结果有两种情况,可以是指定区域中具有可移动体,或者是指定区域中没有可移动体。有几个检测设备对该指定区域进行检测对应的可以得到几个检测结果,而且,这几个检测结果应该是在同一时段内做出的。例如,一个办公室区域中安装有十个不同类型的检测设备,在一个时段内对该办公室是否有人同时或先后进行一次检测,可以得到十个检测结果。
本发明实施例对检测服务器获取多个检测结果的方式不作限定,可以是通过检测服务器与多个检测设备的控制系统连接,每进行一次检测,检测服务器通过该控制系统获取到多个检测结果。
S202:从所述多个检测结果中确定出多个输入参数输入到所述检测模型中。
在本发明实施例中,由该多个检测设备进行一次检测,由于每一个检测设备输出一个检测结果,故检测服务器获取的检测结果可以与检测设备的个数相同,但是指定区域中是否具有可移动体的最终结果只有一种情况,要么具有可移动体,要么没有可移动体。当检测服务器获取到多个检测结果后,需要综合这多个检测结果进行分析处理,从而得出一个最终结果。检测服务器中配置有检测模型,检测模型可以用于对检测结果进行分析,故此,检测服务器可以将获取的多个检测结果确定出多个输入参数输入到所述检测模型中,以便通过该检测模型对这多个检测结果进行分析处理。
检测服务器获取的检测结果可以是指定区域中具有,或者是指定区域中没有可移动体的检测结果,若将该检测结果直接输入到检测模型中,可能会出现检测模型无法识别检测结果的情况,因此,可以将检测结果转换为输入参数输入到所述检测模型中。输入参数可以是检测模型可以识别出的一种参数,是与检测结果相对应的一种参数,可以由检测结果进行转换得出,输入参数可以用数值表示,例如针对指定区域中具有可移动体这一检测结果可以使用参数a来标识,针对指定区域中不具有可移动体这一检测结果可以使用参数b来标识。使用何种参数并不限定,例如可以用a=1表示指定区域中具有可移动体,b=0表示指定区域中没有可移动体,其中“1”和“0”即为输入参数,在一次检测中,检测设备A的检测结果为具有可移动体,检测设备B和检测设备C的检测结果为没有可移动体,则对应的输入参数可以是“1,0,0”。
根据该多个检测结果的个数和该多个输入参数的个数之间的大小关系,本发明实施例提供了多种从多个检测结果中确定出多个输入参数的方式。
一种情况下,该多个检测结果的个数和该多个输入参数的个数相同,且均为N。这种情况下,可以将所述N个检测结果作为N个输入参数输入到所述检测模型中。
在这种情况下,得到这N个检测结果的N个检测设备的类型可以至少有两种可能的组合方式:
一种组合方式可以是检测设备类型的数量为多个,但小于检测设备的个数N,即这N个检测设备中具有相同类型的检测设备,例如例如五个检测设备中包括一个类型a的检测设备,一个类型b的检测设备,三个类型c的检测设备。这种组合方式下,通过检测模型计算确定结果时,由同一个类型的检测设备的检测结果所确定的输入参数可以使用相同的确定权重。而另一种组合方式可以检测设备类型的数量与检测设备的数量相同,即这N个检测设备的类型各不相同。
另一种情况下,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M 小于N,也就是说,检测结果的数量要多于输入参数的数量。这种情况下,需要减少检测结果的数量,由于相同类型的检测设备具有相同的技术缺陷,减少相同类型的检测设备的检测结果对计算出确定结果的影响最小,故可以通过将相同类型的检测设备的检测结果进行聚合,以减少检测结果的数量。
具体的,可以从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M。在一些情况下,Y可以取1。从X个检测结果中确定出Y个检测结果的方式本发明并不限定,例如可以从X个检测结果中任意选出Y个检测结果,也可以按照X个检测结果中有无可移动体的比例,确定出具有相同或相似比例的Y个检测结果。
通过减少检测结果,最终可以从N个检测结果中确定出M个检测结果,从而可以将所述M个检测结果作为M个输入参数输入到所述检测模型中。
S203:所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数。
检测模型中具有针对不同类型检测设备所对应的确定权重,即每个检测设备类型都具有一个对应的确定权重。确定权重可以用于表示一个类型的检测设备所确定的检测结果在该多个类型检测设备所确定的检测结果中所占的比重,确定权重可以是具体的数值,如小数、百分数等形式,确定权重越大说明该确定权重所对应类型的检测设备的检测结果在该多个不同类型检测设备的检测结果中占的比重越大,也即该检测结果对最终得出的确定结果的影响程度越大。
检测模型中的确定权重可以是预设的,也可以是动态调整的。
例如,确定权重可以依据于检测设备的个数进行预先设置,例如,有三个不同类型的检测设备对指定区域是否具有可移动设备进行检测,则可以采用平均值的方式,将每个类型的检测设备所对应的确定权重设置为1/3。
确定权重也可以通过对多次历史检测结果的分析计算得出,例如,有三个不同类型的检测设备,检测设备A、检测设备B和检测设备C对指定区域是否具有可移动设备进行检测,每进行一次检测,将三个不同类型的检测设备的检测结果与实际结果进行比较,以可移动体为人为例,检测设备A的检测结果是没有人、检测设备B和检测设备C的检测结果是有人,实际结果是没有人,则说明检测设备A的检测结果正确,检测设备B和检测设备C的检测结果错误,并将每个设备的检测结果的正确和错误情况进行记录,以 此类推,可以进行多次检测,依据每次检测的记录可以统计出检测设备A中检测结果正确的次数,检测设备B中检测结果正确的次数以及检测设备C中检测结果正确的次数,将三个检测设备中检测结果正确的次数相加可以得到一个总数值,每个检测设备的检测结果所对应的确定权重可以由该检测设备检测结果正确的次数除以该总数值得出。
例如,在100次检测中,检测设备A中检测结果正确的次数为80,检测设备B中检测结果正确的次数为50以及检测设备C中检测结果正确的次数为70,将三个检测设备中检测结果正确的次数相加可以得到一个总数值200,则可以计算出检测设备A所对应的确定权重为80/200=2/5,检测设备B所对应的确定权重为50/200=1/4,检测设备C所对应的确定权重为70/200=7/20。而且,这三个不同类型的检测设备的确定权重还可以根据分析计算的样本数增加而随之动态调整。
需要注意的是,在一种可选的方案中,并不需要每次检测时,都将各个检测设备的检测结果作为动态调整确定权重的依据。为了提高计算效率。可以将确定结果与实际结果不符时各个检测设备的检测结果作为调整确定权重的依据,若确定结果与实际结果相符,证明使用目前的确定权重可以得到正确的确定结果,故可以保持现有确定权重不变。
在S202中检测服务器从获取的多个检测结果中确定出多个输入参数输入到检测模型中,检测模型可以根据该多个检测设备类型各自对应的确定权重,对多个输入参数进行处理从而得到一个输出参数,输出参数可以是一个具体的数值,可以由检测模型根据输入参数和确定权重计算得出。一种可行的计算输出参数的方式,可以是将多个输入参数与其对应的确定权重相乘,得出多个乘积值,再将这多个乘积值相加从而得出一个输出参数,以三个输入参数为例,不同类型的检测设备A、检测设备B和检测设备C各自对应的确定权重分别为2/5、2/5和1/5,在一次检测中,检测设备A的检测结果为具有可移动体,其对应的输入参数可以设置为“1”,检测设备B和检测设备C的检测结果为没有可移动体,其对应的输入参数可以设置为“0”,则输出参数=2/5*1+2/5*0+1/5*0=2/5。
S204:根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
确定结果可以是针对一次检测处理得到的最终结果,可以理解为是检测服务器将多个检测结果综合分析后得出的一个最终结果。该确定结果有两种情况,可以是指定区域中具有可移动体,或者是指定区域中没有可移动体。
在S203中可以得出一个输出参数,本发明实施例,需要依据该输出参数最终确定出指定区域中是否具有可移动体,因此,需要对得出的输出参数进行分析得出一个确定结 果。具体的确定方式本发明并不限定,例如,可以根据输入参数中针对指定区域中具有可移动体的赋值作为得到确定结果的依据,若输出参数与该赋值比较接近,那么可以认为根据该输出参数得出的确定结果为指定区域中具有可移动体,反之则得出指定区域中不具有可移动体的确定结果。
为了方便计算,本发明实施例给出了一种根据输出参数得到确定结果的具体方式。输出参数可以是一个具体的数值,对于确定结果的两种情况也可以设置相对应的数据,例如,可以用第一数据来表示指定区域中具有可移动体,可以用第二数据来表示指定区域中没有可移动体,通过判断该输出参数与第一数据和第二数据的接近程度,从而确定出指定区域中是否具有可移动体。具体操作如下:
判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度;若所述输出数据相对于第二数据更接近第一数据,得出所述指定区域中具有可移动体的确定结果;若所述输出数据相对于第一数据更接近第二数据,得出所述指定区域中没有可移动体的确定结果。
接近程度可以用于表示输出数据与第一数据和第二数据的相关性,第一数据可以用于标识具有可移动体的数据,第二数据可以用于标识没有可移动体的数据,当输出数据与第一数据的接近程度相对于第二数据更高时,说明输出数据与第一数据的相关性更高,那么指定区域中具有可移动体的可能性较高,即确定结果可以是所述指定区域中具有可移动体;当输出数据与第二数据的接近程度相对于第一数据更高时,说明输出数据与第二数据的相关性更高,那么指定区域中没有可移动体的可能性较高,即确定结果可以是所述指定区域中没有可移动体。
其中,第一数据和第二数据的具体数值可以依据于输出参数的数值范围进行设置,其中,输出参数的数值范围可以根据输入参数得出。例如,输入参数采用“100”和“0”两种数值表示,“100”表示检测结果为具有可移动体,“0”表示检测结果没有可移动体,依据S203计算输出参数的方式,可以确定出输出参数可以是处于0-100范围内的数值,故此,可以将第一数据设置为“100”,第二数据设置为“0”。
本发明实施例中对判断接近程度的方式不作限定,一种可行的方式可以是采用差值来计算接近程度,分别计算输出参数与第一数据和输出参数与第二数据的差值,差值越小则表示接近程度越高。例如,第一数据为100,第二数据为0,输出数据为70,该输出数据与第一数据的差值为30,该输出数据与第二数据的差值为70,该输出数据与第一数据的差值比与第二数据的差值小,则表示该输出数据与第一数据的接近程度更高,即 该输出数据与第一数据的相关性更高,第一数据可以是用于标识“具有可移动体”这一结果的数据,那么指定区域中具有可移动体的可能性较高,所以可以得出确定结果为指定区域中具有可移动体。另一种可行的方式可以是根据检测精度需求对接近程度进行调整,例如在检测精度较高时,输出参数与第一数据之间的差值需要非常小才能有较高的接近程度,例如,第一数据为100,第二数据为0,输出数据为70,设置输出数据需要与第一数据差值最大等于10才能认定输出数据相对更接近第一数据,这种检测精度下,输出数据为70的情况将被判定为与第二数据接近程度较高,由于第二数据可以是用于标识“不具有可移动体”这一结果的数据,那么指定区域中不具有可移动体的可能性较高,所以可以得出确定结果为指定区域中不具有可移动体。
由上述技术方案可以看出,检测服务器中配置有针对不同类型检测设备所对应的确定权重的检测模型,当获取到多个不同类型的检测设备针对指定区域中是否具有可移动体的多个检测结果后,可以从该多个检测结果中确定出多个输入参数输入到所述检测模型中,由于检测结果来自不同类型的检测设备,故根据输出检测结果的检测设备的类型确定出该检测设备类型对应的确定权重,检测模型根据检测设备类型所对应的确定权重以及根据检测设备的检测结果所确定的多个输入参数可以得到输出参数,通过对所述输出参数的分析可以得到针对所述指定区域中是否具有可移动体的确定结果,可见,对指定区域的确定结果需要参考多个不同类型检测设备的检测结果,即使其中某些类型的检测设备的检测结果出现误判,也有可能根据其他类型的检测设备所输出的正确检测结果而得出正确的确定结果,从而尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高了检测的准确性。
考虑到可能会出现确定结果与实际结果不符合的情况,例如,得出的确定结果是指定区域中没有可移动体,但是实际结果是该指定区域中具有可移动体。针对上述情况,可以在得出确定结果后将其与实际结果进行比较,若确定结果与实际结果不同,说明该确定结果出现错误。为了进一步提高检测的准确性,对于确定结果出现错误的情况,可以增加对确定结果的反馈操作。具体的,在S204之后,还包括:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,判断所述N个检测结果中是否有与所述实际结果相同;若具有与所述实际结果相同的检测结果,相应增加所述检测模型中所述与实际结果相同的检测结果的检测设备类型所对应的确定权重。
这里的实际结果可以理解为得出确定结果的时段中,该指定区域中的是否有可移动体的真实情况。实际结果可以通过验证得到,该验证可以由人工验证,或者是借助于机器验证,例如,对一间办公室中是否有人进行检测,可以通过安排工作人员到该办公室进行检查确定出该办公室实际是否有人,或者是通过监控系统监控该办公室的实际情况。
实际结果也可以依据于指定区域中的相关操作得出,例如,对一间办公室是否有人进行检测,确定结果为办公室中没有人,该办公室的照明控制系统可以自动将照明设备关闭,当办公室中实际有人,如工作人员,关闭照明设备显然干扰到该工作人员的正常工作,该工作人员可以通过控制开关将照明设备重新打开,故若检测到关闭的照明设备很快被重新打开这一行为,则可以反映出实际结果是办公室有人,通过照明设备关闭和照明设备重新打开这两种行为可以判断出确定结果与实际结果不同。
由前述介绍可知,确定结果是由检测服务器对多个检测结果进行综合分析后得出的,当确定结果与实际结果不同时,说明多个检测结果中有出现错误的检测结果,并且这些错误的检测结果对确定结果的影响较大,从而导致最终得出的确定结果错误。
确定权重是反映检测结果对确定结果影响度的重要因素,当得到错误检测结果的检测设备类型相比于得到正确检测结果的检测设备类型所对应的确定权重较高时,也即检测结果中得到正确检测结果的检测设备类型相比于得到错误检测结果的检测设备类型所对应的确定权重较低时,则错误的检测结果对确定结果的影响更大,从而导致确定结果与实际结果不符。例如,有三个不同类型的检测设备,检测设备A、检测设备B和检测设备C,其对应的确定权重均为1/3,在一次检测中,通过与实际结果比较,检测设备A的检测结果正确,检测设备B和检测设备C的检测结果错误,可知检测结果中正确检测结果的检测设备类型所对应的确定权重为1/3,错误检测结果的检测设备类型所对应的确定权重为2/3,相比之下,由于得到正确检测结果检测设备A占的确定权重较低,从而导致最终得到的确定结果与检测设备B和检测设备C的检测结果一致,即确定结果与实际结果不符。
由此可知,得到正确检测结果的检测设备类型相比于得到错误检测结果的检测设备类型所占的确定权重较低,是导致确定结果出现错误的主要因素。若检测设备类型所对应的确定权重未发生改变,当下一次出现类似的检测情景时,确定结果依旧会出现与实际结果不符的错误。因此,当确定结果与实际结果不同时,可以通过增加得到与实际结果相同检测结果的检测设备类型所对应的确定权重,来避免上述情况的发生。当得到与实际结果相同检测结果的检测设备类型所对应的确定权重增加后,则该类型的检测设备 所得到的检测结果可以在确定结果中起到更大的作用,从而使得确定结果与该检测结果相同的可能性较大,即确定结果与实际结果相符的可能性更高。
需要说明的是,由于一个类型的检测设备所对应的确定权重可以由该检测设备的检测结果在该多个不同类型的检测设备的检测结果中所占的比重得出,故该多个不同类型的检测设备所对应的确定权重相加一般为一个固定数值,因此,当增加检测模型中与实际结果相同的检测结果所对应的确定权重,相应的检测模型中与实际结果不同的检测结果所对应的确定权重会减少,为了方便说明,以下例子中,得到确定结果的过程中,均以采用差值来计算接近程度。
例如,不同类型的检测设备A、检测设备B和检测设备C所对应的确定权重均为1/3,所有确定权重相加为1,检测设备A的检测结果与实际结果相同,检测设备B和检测设备C的检测结果与实际结果不同,此时,得到与实际结果相同检测结果的检测设备A所对应的确定权重为1/3,得到与实际结果不同检测结果的检测设备B和检测设备C所对应的确定权重之和为2/3,最终结果与检测设备B和检测设备C的检测结果相同,即最终结果与实际结果不符,为了使得当再次出现上述检测结果时,最终得出的确定结果能够与实际结果相符,而不是重新得到与实际结果不符的确定结果,故需要增加检测设备A所对应的确定权重,例如将其调整为3/5,则检测设备B和检测设备C所对应的确定权重会相应的减少为2/5。从而当再次出现上述检测结果时,最终得出的确定结果与实际结果相符。
由上述介绍可知,当确定结果与实际结果不同时,可以通过增加检测模型中得到与实际结果相同检测结果的检测设备类型所对应的确定权重,实现对检测模型的优化,从而进一步提升检测的准确性。
接下来将对增加确定权重的具体方式展开介绍,在本发明实施例中,当确定结果与实际结果不符时,可以通过建立数据序列的方式来表示该多个不同类型检测设备的检测结果的实际情况,每一个数据序列包括该多个数据位,每一个数据位对应一个类型的检测设备。通过数据序列中记录的参数,作为动态调整检测模型中确定权重的依据。
本发明实施例对数据位与检测设备的对应顺序不作限定,一个数据序列包括M个数据位可以按照检测设备的顺序进行排列,例如,M个不同类型的检测设备依次为L1-LM,对应的一个数据序列包括M个数据位,第一个数据位可以对应于第一个检测设备L1,第二个数据位可以对应于第二个检测设备L2,以此类推,第M个数据位可以对应于第M个检测设备LM。
一个数据序列包括M个数据位也可以随机排列,例如,有三个不同类型的检测设备依次为L1-L3,第一个数据位可以对应于第一个检测设备L1,第二个数据位对应于第三个检测设备L3,第三个数据位对应于第二个检测设备L2。
一般来说,当确定结果与实际结果相同时,可以认为当前使用的确定权重配置的相对合理,可以得出正确的确定结果,当确定结果与实际结果不同时,可以认为当前使用的确定权重配置的可能并不合理,导致错误的检测结果对整体的影响较大,最终得出了错误的确定结果。故可以在出现错误的确定结果时,通过建立一个新的数据序列即目标序列,将本次各个检测设备得出的检测结果对应填入到该目标序列中,并依据该目标序列实现对检测结果的确定权重的调整,如图3所示,具体操作如下:
S301:根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出所述与实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子。
在每一个数据位中可以通过填充有效因子或者无效因子来表示检测设备的检测结果的正确性。有效因子可以用于表示检测设备的检测结果与实际结果相同,无效因子可以用于表示检测设备的检测结果与实际结果不同,有效因子和无效因子的形式可以有多种,可以是数字、字母或者文字等,本发明实施例对其形式不作限定。例如,数字代表有效因子,字母代表无效因子,则在数据位中填充的因子是数字形式的即为有效因子,在数据位中填充的因子是字母形式的即为无效因子。又或者,小于100的数字代表有效因子,大于或等于100的数字代表无效因子,则在数据位中填充的因子是小于100的数字即为有效因子,在数据位中填充的因子是大于或等于100的数字即为无效因子。为了后续方便介绍,接下来均以有效因子为“1”,无效因子为“0”为例介绍,即“1”表示检测结果与实际结果相同,“0”表示检测结果与实际结果不同。
检测模型中可以具有多个数据序列,该多个数据序列中的数据可以是预先设置的试验数据,例如通过预置出各类较难识别的场景,以达到确定结果与实际结果不符的情况,从而将确定结果与实际结果不符时的各个检测设备得到的检测结果作为试验数据填入数据序列中。该多个数据序列中的数据也可以是随机填入的数据,并以此作为之后循环迭代计算确定权重的基础。该多个数据序列中的数据也可以是在正常检测过程中,出现确定结果与实际结果不符时的各个检测设备得到的检测结果所对应的数据。例如,不同类型的检测设备A、检测设备B和检测设备C对指定区域是否有人进行检测,检测设备A的检测结果为有人、检测设备B和检测设备C的检测结果均为无人,根据这三个检测结 果得到的确定结果为指定区域中无人,而该指定区域的实际结果是有人,可见确定结果与实际结果不符。其中,检测设备A的检测结果与实际结果相同,检测设备B和检测设备C的检测结果与实际结果不同,依据这三个检测设备的检测结果可以建立一个数据序列“1,0,0”。
目标序列可以反映出确定结果与实际结果不同时,与实际结果相比多个检测结果的正确和错误情况。由于确定权重的调整与有效因子的数量相关,故目标序列的建立可以用于增加与实际结果相同的检测结果的检测设备所对应的数据位的有效因子个数。例如,不同类型的检测设备A、检测设备B和检测设备C对指定区域是否有人进行检测,检测设备A的检测结果为无人、检测设备B和检测设备C的检测结果均为有人,检测设备A所对应的确定权重高于检测设备B和检测设备C所对应的确定权重,由此得出的确定结果为无人,但是实际结果为有人,导致最终得到的确定结果与实际结果不符,由于检测设备A的检测结果与实际结果不同,检测设备B和检测设备C的检测结果与实际结果相同,则建立的目标序列为“0,1,1”,以希望一定程度上提高检测设备B的类型和检测设备C的类型对应的确定权重。
又例如,不同类型的检测设备A、检测设备B和检测设备C对指定区域是否有人进行检测,检测设备A的检测结果为有人、检测设备B和检测设备C的检测结果均为无人,检测设备A所对应的确定权重高于检测设备B和检测设备C所对应的确定权重,由此得出的确定结果为有人,但是实际结果为无人,导致最终得到的确定结果与实际结果不符,由于检测设备A的检测结果与实际结果不同,检测设备B和检测设备C的检测结果与实际结果相同,则建立的目标序列为“0,1,1”,以希望一定程度上提高检测设备B的类型和检测设备C的类型对应的确定权重。
S302:根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出所述与实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的总数,调整所述与实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出所述与实际结果相同的检测结果的检测设备中的任意一个检测设备。
以第一设备为例,一种可行的计算该第一设备所对应的确定权重方式可以是,将该第一设备所对应数据位中填充的有效因子的总数,与检测模型的全部数据序列中所填充 的有效因子的总数相除计算得出,例如表1所示:
第一设备 第二设备 第三设备
1 0 0
0 1 0
0 0 1
1 0 0
表1
表1中,“1”代表有效因子,“0”代表无效因子,表1中第一行代表不同类型的检测设备,分别为第一设备、第二设备和第三设备,第二行、第三行和第四行为检测模型中具有的三个数据序列,第五行是建立的目标序列,每一列数据是与检测设备对应的数据,第一列数据是第一设备对应的数据。
若目标序列具体为第五行所示,可以在几种可能的情况下:
一种可能的情况,针对指定区域,当第一设备的检测结果为有人,第二设备和第三设备的检测结果为无人时,若综合这三个检测结果得到的确定结果为无人,但是实际结果为有人时,相当于得到与实际结果相符的检测结果的第一设备在这三个检测设备中对确定结果的影响较小,需要调高第一设备对应的确定权重,从而录入的目标序列,即第5行所示数据序列可以为“1,0,0”。
还有一种可能的情况,针对指定区域,当第一设备的检测结果为无人,第二设备和第三设备的检测结果为有人时,若综合这三个检测结果得到的确定结果为有人,但是实际结果为无人时,相当于得到与实际结果相符的检测结果的第一设备在这三个检测设备中对确定结果的影响较小,需要调高第一设备对应的确定权重,从而录入的目标序列,即第5行所示数据序列可以为“1,0,0”。
由表1可知,检测模型的全部数据序列中所填充的有效因子的总数为4个,全部数据序列中第一设备所对应数据位中填充的有效因子的总数为2个,则该第一设备所对应的确定权重从原本的1/3调高为2/4=1/2。
上述过程中,以第一设备(与实际结果相同的检测结果的检测设备)为例,对其确定权重的调整展开的介绍。对于与实际结果不同的检测结果的检测设备,在该检测设备的数据位上填充的是无效因子,其数据位上填充的有效因子的总数未发生变换,但是由于与实际结果相同的检测结果的检测设备的数据位上填充的是有效因子,使得全部数据序列中填充的有效因子的总数发生了变化,所以与实际结果不同的检测结果的检测设备 类型所对应的确定权重也会发生相应的变化,其确定权重的计算方法与第一设备所对应的确定权重的计算方法类似,在此不再赘述。
例如表2所示:
检测设备A 检测设备B 检测设备C
1 0 0
0 1 0
0 0 1
0 1 1
表2
表2中,“1”代表有效因子,“0”代表无效因子,表中第一行代表不同类型的检测设备,分别为检测设备A、检测设备B和检测设备C,第二行、第三行和第四行为检测模型中具有的三个数据序列,第五行是建立的目标序列,每一列数据是与检测设备对应的数据,第一列数据是检测设备A对应的数据,第二列数据是检测设备B对应的数据,第三列数据是检测设备C对应的数据。
若目标序列具体为第五行所示,可以在几种可能的情况下:
一种可能的情况,针对指定区域,当检测设备A的检测结果为有人,检测设备B和检测设备C的检测结果为无人时,若综合这三个检测结果得到的确定结果为有人,但是实际结果为无人时,相当于得到与实际结果相符的检测结果的检测设备B和检测设备C在这三个检测设备中对确定结果的影响较小,需要调高检测设备B和检测设备C对应的确定权重,从而录入的目标序列,即第5行所示数据序列可以为“0,1,1”。
还有一种可能的情况,针对指定区域,当检测设备A的检测结果为无人,检测设备B和检测设备C的检测结果为有人时,若综合这三个检测结果得到的确定结果为无人,但是实际结果为有人时,相当于得到与实际结果相符的检测结果的检测设备B和检测设备C在这三个检测设备中对确定结果的影响较小,需要调高检测设备B和检测设备C对应的确定权重,从而录入的目标序列,即第5行所示数据序列可以为“0,1,1”。
由表中可以看出检测模型的全部数据序列中所填充的有效因子的总数为5个,在全部数据序列中,检测设备A所对应数据位中填充的有效因子的总数为1个,则检测设备A所对应的确定权重为1/5,检测设备B所对应数据位中填充的有效因子的总数为2个,则检测设备B所对应的确定权重从1/3调高为2/5,检测设备C所对应数据位中填充的有效因子的总数为2个,则检测设备C所对应的确定权重从1/3调高为2/5。
需要说明的是,本发明实施例中确定权重的调整主要依据有效因子和无效因子的个数进行,也即有效因子和无效因子的形式对确定权重的调整并无影响,因此,即使对于同一个目标序列,填充的有效因子也可以有多种形式,例如,M=10,10个不同类型的检测设备中有6个检测设备的检测结果与实际结果相同,则建立的一个目标序列中有6个有效因子,这6个有效因子的形式可以不相同。
在上述过程中,对于确定结果与实际结果不符的情况,通过增加与实际结果相同的检测结果的检测设备的有效因子的个数,以提高与实际结果相同的检测结果的检测设备类型所对应的确定权重,使得该检测设备的检测结果可以对确定结果具有更高的影响度,从而进一步提高确定结果的准确性。
接下来从交互端的角度来描述本发明的技术方案,这里所述的交互端可以是具有交互功能的智能设备,例如手机、PAD、个人计算机等。该交互端具有可以和检测服务器建立数据连接的功能,从而用户通过交互端可以获知针对指定区域的确定结果,并可以在与实际结果不同的情况下,通过交互端向检测服务器指示该确定结果错误。
图4为本发明实施例提供的一种检测方法的方法流程图,应用于交互端,所述方法包括:
S401:获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的。
本步骤中各个特征的说明可以参见图2所对应实施例中的相关描述,这里不再赘述。
该确定结果可以是由检测服务器发送的,也可以是其他设备发送的,本发明对此并不限定。通过获取的通过综合各个不同类型检测设备的检测结果所得出的确定结果,交互端可以明确该确定结果指示该指定区域中是否具有可移动体。
S402:若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
实际结果可以是通过交互端的使用者观察获得,也可以是通过该指定区域中产生的相关操作所确定,若确定结果为无可移动体,该相关操作例如可以是在电源被关闭后突然又从该指定区域中被打开了。
当确定出该确定结果与该指定区域中实际结果不同时,交互端可以生成调整指令并 输出,例如向检测服务器发送该调整指令,以告知检测服务器针对该指定区域做出的确定结果是错误的。从而检测服务器可以根据检测结果中与实际结果之间的关系,调整不同检测设备类型所对应的确定权重,以提高与做出与实际结果相同的检测设备的类型的确定权重。具体如何调整可以参见图2、图3所对应实施例中相关描述,这里不再赘述。
需要注意的是,若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,也就是确定结果是正确的情况下,交互端可以根据所述确定结果控制所述指定区域中的可控设备。
这里所述的可控设备可以指可以远程控制的设备,主要可以包括该指定区域提供服务性功能的设备,例如可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
在确定结果为无可移动体的情况下,表示该指定区域中不再需要可控设备提供服务,故交互段可以关闭该指定区域中的可控设备,例如照明、空调等。
在确定结果为具有可移动体的情况下,表示该指定区域中需要可控设备继续提供服务,故交互段可以控制该指定区域中的可控设备处于服务状态,例如将关闭的照明、空调等打开。
可见,通过交互端对确定结果的核实,可以有效减少错误的确定结果对指定区域所带来的不良影响,并通过对核实结果的反馈,为检测服务器调整计算确定结果的方式提供依据,为提高确定结果的准确性做出了贡献。
图5为本发明实施例提供的一种检测服务器,所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器包括获取单元501和确定单元502:
所述获取单元501,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;
所述确定单元502,用于从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;
所述确定单元502还用于利用所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;
所述确定单元502还用于根据所述输出参数得到针对所述指定区域中是否具有可移 动体的确定结果。
可选的,所述检测服务器还包括判断单元和增加单元:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述判断单元,所述判断单元,用于判断所述多个检测结果中是否有与所述实际结果相同;
若具有与所述实际结果相同的检测结果,触发所述增加单元,所述增加单元用于相应增加所述检测模型中所述与实际结果相同的检测结果的检测设备类型所对应的确定权重。
可选的,所述检测模型中具有多个数据序列,每一个数据序列包括M个数据位,每一个数据位对应一个类型的检测设备,所述增加单元还用于根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出所述与实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子;
根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出所述与实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的总数,调整所述与实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出所述与实际结果相同的检测结果的检测设备中的任意一个检测设备。
可选的,所述确定单元包括判断子单元和确定子单元:
所述判断子单元,用于判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度,所述第一数据为用于标识具有可移动体的数据,所述第二数据为用于标识没有可移动体的数据;
若所述输出数据相对于第二数据更接近第一数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中具有可移动体的确定结果;
若所述输出数据相对于第一数据更接近第二数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中没有可移动体的确定结果。
可选的,所述检测设备的类型至少包括人体占用传感器、定位传感器和噪声传感器。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为N个,所述确定单元还用于将所述N个检测结果作为N个输入参数输入到所述检测模型中。
可选的,所述N个检测设备的类型各不相同。
可选的,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M小于N,所述N个检测设备具有M个不同类型,所述确定单元还用于从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M;
将所述M个检测结果作为M个输入参数输入到所述检测模型中。
可见,检测服务器中配置有针对不同类型检测设备所对应的确定权重的检测模型,当获取到多个不同类型的检测设备针对指定区域中是否具有可移动体的多个检测结果后,可以从该多个检测结果中确定出多个输入参数输入到所述检测模型中,由于检测结果来自不同类型的检测设备,故根据输出检测结果的检测设备的类型确定出该检测设备类型对应的确定权重,检测模型根据检测设备类型所对应的确定权重以及根据检测设备的检测结果所确定的多个输入参数可以得到输出参数,通过对所述输出参数的分析可以得到针对所述指定区域中是否具有可移动体的确定结果,可见,对指定区域的确定结果需要参考多个不同类型检测设备的检测结果,即使其中某些类型的检测设备的检测结果出现误判,也有可能根据其他类型的检测设备所输出的正确检测结果而得出正确的确定结果,从而尽可能的减少单一类型检测设备由于技术局限性出现误判对用户体验所产生的不良影响,并有效的提高了检测的准确性。
图6为本发明实施例提供的一种交互端,所述交互端包括获取单元601和输出单元602:
所述获取单元601,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;
若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述输出单元602,所述输出单元602,用于输出调整指令,所述调整指令用于指示所述确定结果是错误的。
可选的,还包括控制单元:
若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,触发所述控制单元,所述控制单元,用于根据所述确定结果控制所述指定区域中的可控设备。
可选的,所述可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
可见,通过交互端对确定结果的核实,可以有效减少错误的确定结果对指定区域所带来的不良影响,并通过对核实结果的反馈,为检测服务器调整计算确定结果的方式提供依据,为提高确定结果的准确性做出了贡献。
图7为本发明实施例提供的一种检测系统,所述检测系统包括检测服务器701和交互端702:
所述检测服务器701中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器701,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
所述交互端702,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
本领域普通技术人员可以理解:实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成,前述程序可以存储于一计算机可读取存储介质中,该程序在执行时,执行包括上述方法实施例的步骤;而前述的存储介质可以是下述介质中的至少一种:只读存储器(英文:read-only memory,缩写:ROM)、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
需要说明的是,本说明书中的各个实施例均采用递进的方式描述,各个实施例之间相同相似的部分互相参见即可,每个实施例重点说明的都是与其他实施例的不同之处。尤其,对于设备及系统实施例而言,由于其基本相似于方法实施例,所以描述得比较简单,相关之处参见方法实施例的部分说明即可。以上所描述的设备及系统实施例仅仅是示意性的,其中作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单 元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动的情况下,即可以理解并实施。
以上所述,仅为本发明较佳的具体实施方式,但本发明的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本发明揭露的技术范围内,可轻易想到的变化或替换,都应涵盖在本发明的保护范围之内。因此,本发明的保护范围应该以权利要求的保护范围为准。

Claims (23)

  1. 一种检测方法,其特征在于,应用于检测服务器,所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述方法包括:
    获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;
    从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;
    所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;
    根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
  2. 根据权利要求1所述的方法,其特征在于,在所述根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果之后,还包括:
    若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,判断所述多个检测结果中是否有与所述实际结果相同;
    若具有与所述实际结果相同的检测结果,相应增加所述检测模型中与所述实际结果相同的检测结果的检测设备类型所对应的确定权重。
  3. 根据权利要求2所述的方法,其特征在于,所述检测模型中具有多个数据序列,每一个数据序列包括M个数据位,每一个数据位对应一个类型的检测设备,所述相应增加所述检测模型中与所述实际结果相同的检测结果的检测设备类型所对应的确定权重,包括:
    根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出与所述实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子;
    根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出与所述实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的总数,调整与所述实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出与所述实际结果相同的检测结果的检测设备中的任意一个检测设备。
  4. 根据权利要求1至3任意一项所述的方法,其特征在于,所述根据所述输出参数 得到针对所述指定区域中是否具有可移动体的确定结果,包括:
    判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度,所述第一数据为用于标识具有可移动体的数据,所述第二数据为用于标识没有可移动体的数据;
    若所述输出数据相对于第二数据更接近第一数据,得出所述指定区域中具有可移动体的确定结果;
    若所述输出数据相对于第一数据更接近第二数据,得出所述指定区域中没有可移动体的确定结果。
  5. 根据权利要求1所述的方法,其特征在于,所述检测设备的类型至少包括人体占用传感器、定位传感器和噪声传感器。
  6. 根据权利要求1所述的方法,其特征在于,所述多个检测结果具体为N个,所述多个输入参数具体为N个,所述从所述多个检测结果中确定出多个输入参数输入到所述检测模型中,包括:
    将所述N个检测结果作为N个输入参数输入到所述检测模型中。
  7. 根据权利要求6所述的方法,其特征在于,所述N个检测设备的类型各不相同。
  8. 根据权利要求1所述的方法,其特征在于,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M小于N,所述N个检测设备具有M个不同类型,所述从所述多个检测结果确定中出多个输入参数输入到所述检测模型中,包括:
    从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M;
    将所述M个检测结果作为M个输入参数输入到所述检测模型中。
  9. 一种检测方法,其特征在于,应用于交互端,所述方法包括:
    获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;
    若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
  10. 根据权利要求9所述的方法,其特征在于,还包括:若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,根据所述确定结果控制所述指定区域中的可 控设备。
  11. 根据权利要求10所述的方法,其特征在于,所述可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
  12. 一种检测服务器,其特征在于,所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器包括获取单元和确定单元:
    所述获取单元,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;
    所述确定单元,用于从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;
    所述确定单元还用于利用所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;
    所述确定单元还用于根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果。
  13. 根据权利要求12所述的检测服务器,其特征在于,所述检测服务器还包括判断单元和增加单元:
    若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述判断单元,所述判断单元,用于判断所述多个检测结果中是否有与所述实际结果相同;
    若具有与所述实际结果相同的检测结果,触发所述增加单元,所述增加单元用于相应增加所述检测模型中与所述实际结果相同的检测结果的检测设备类型所对应的确定权重。
  14. 根据权利要求13所述的检测服务器,其特征在于,所述检测模型中具有多个数据序列,每一个数据序列包括M个数据位,每一个数据位对应一个类型的检测设备,所述增加单元还用于根据所述判断所述多个检测结果中是否有与所述实际结果相同的判断结果建立目标序列,所述目标序列为一个数据序列,所述目标序列中,输出与所述实际结果相同的检测结果的检测设备所对应的数据位中填充有效因子,其他数据位中填充无效因子;
    根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中输出与所述实际结果相同的检测结果的检测设备所对应数据位中填充的有效因子的 总数,调整与所述实际结果相同的检测结果的检测设备类型所对应的确定权重;其中,针对第一设备,根据所述检测模型的全部数据序列中所填充的有效因子的总数,以及全部数据序列中第一设备所对应数据位中填充的有效因子的总数,确定出所述第一设备所对应的确定权重,所述第一设备为输出与所述实际结果相同的检测结果的检测设备中的任意一个检测设备。
  15. 根据权利要求12至14任意一项所述的检测服务器,其特征在于,所述确定单元包括判断子单元和确定子单元:
    所述判断子单元,用于判断所述检测模型的输出数据与第一数据和第二数据之间的接近程度,所述第一数据为用于标识具有可移动体的数据,所述第二数据为用于标识没有可移动体的数据;
    若所述输出数据相对于第二数据更接近第一数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中具有可移动体的确定结果;
    若所述输出数据相对于第一数据更接近第二数据,触发所述确定子单元,所述确定子单元,用于得出所述指定区域中没有可移动体的确定结果。
  16. 根据权利要求12所述的检测服务器,其特征在于,所述检测设备的类型至少包括人体占用传感器、定位传感器和噪声传感器。
  17. 根据权利要求12所述的检测服务器,其特征在于,所述多个检测结果具体为N个,所述多个输入参数具体为N个,所述确定单元还用于将所述N个检测结果作为N个输入参数输入到所述检测模型中。
  18. 根据权利要求17所述的检测服务器,其特征在于,所述N个检测设备的类型各不相同。
  19. 根据权利要求12所述的检测服务器,其特征在于,所述多个检测结果具体为N个,所述多个输入参数具体为M个,M小于N,所述N个检测设备具有M个不同类型,所述确定单元还用于从均为一个类型的X个检测设备的检测结果中确定出Y个作为这个类型的检测设备所对应的检测结果,以此从所述N个检测结果中确定出M个检测结果,X大于Y,X小于N,Y小于M;
    将所述M个检测结果作为M个输入参数输入到所述检测模型中。
  20. 一种交互端,其特征在于,所述交互端包括获取单元和输出单元:
    所述获取单元,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设 备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;
    若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,触发所述输出单元,所述输出单元,用于输出调整指令,所述调整指令用于指示所述确定结果是错误的。
  21. 根据权利要求20所述的交互端,其特征在于,还包括控制单元:
    若所述确定结果与所述指定区域中是否具有可移动体的实际结果相同,触发所述控制单元,所述控制单元,用于根据所述确定结果控制所述指定区域中的可控设备。
  22. 根据权利要求21所述的交互端,其特征在于,所述可控设备包括照明设备、空调、显示设备和投影设备中的任意一个或多个的组合。
  23. 一种检测系统,其特征在于,所述检测系统包括检测服务器和交互端:
    所述检测服务器中配置有检测模型,所述检测模型中具有针对不同类型检测设备所对应的确定权重,所述检测服务器,用于获取多个检测设备针对指定区域中是否具有可移动体的多个检测结果,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果;从所述多个检测结果中确定出多个输入参数输入到所述检测模型中;所述检测模型根据所述多个检测设备所具有类型对应的确定权重以及所述多个输入参数得到输出参数;根据所述输出参数得到针对所述指定区域中是否具有可移动体的确定结果;
    所述交互端,用于获取针对指定区域中是否具有可移动体的确定结果,所述指定区域中具有多个检测设备,在具有多个不同类型的所述多个检测设备中,每一个检测设备输出一个检测结果,所述确定结果为根据所述多个检测设备针对指定区域中是否具有可移动体的多个检测结果确定得出的;若所述确定结果与所述指定区域中是否具有可移动体的实际结果不同,输出调整指令,所述调整指令用于指示所述确定结果是错误的。
PCT/CN2017/117065 2016-12-29 2017-12-19 一种检测方法、相关装置和系统 Ceased WO2018121330A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611247432.8A CN108254801A (zh) 2016-12-29 2016-12-29 一种检测方法、相关装置和系统
CN201611247432.8 2016-12-29

Publications (1)

Publication Number Publication Date
WO2018121330A1 true WO2018121330A1 (zh) 2018-07-05

Family

ID=62710078

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/117065 Ceased WO2018121330A1 (zh) 2016-12-29 2017-12-19 一种检测方法、相关装置和系统

Country Status (3)

Country Link
CN (1) CN108254801A (zh)
TW (1) TW201823932A (zh)
WO (1) WO2018121330A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109143043A (zh) * 2018-09-30 2019-01-04 深圳市盈科互动科技有限公司 控制器io输入防抖检测方法、系统、装置及存储介质
CN112051621A (zh) * 2019-08-15 2020-12-08 河南紫联物联网技术有限公司 一种房间有人无人状态的判定方法与装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101474481A (zh) * 2009-01-12 2009-07-08 北京科技大学 情感机器人系统
CN201654862U (zh) * 2010-01-14 2010-11-24 李元金 校车行驶记录监控仪
US20130196649A1 (en) * 2012-01-27 2013-08-01 Qualcomm Incorporated Mobile device to detect unexpected behaviour
CN104658140A (zh) * 2013-11-21 2015-05-27 青岛唐鹏钢结构工程有限公司 一种应用于防盗门的多路检测智能防盗报警器
CN106019267A (zh) * 2015-03-31 2016-10-12 新明和工业株式会社 运动物体检测装置

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002037827A2 (en) * 2000-10-30 2002-05-10 Naval Postgraduate School Method and apparatus for motion tracking of an articulated rigid body
KR100866485B1 (ko) * 2006-08-22 2008-11-03 삼성전자주식회사 다접점 위치 변화 감지 장치, 방법, 및 이를 이용한 모바일기기
CN101558367A (zh) * 2006-12-05 2009-10-14 索尼爱立信移动通讯有限公司 用于检测物体移动的方法和系统
JP5526100B2 (ja) * 2011-10-18 2014-06-18 Eizo株式会社 動体判別装置、表示装置及び動体判別方法
CN105335452A (zh) * 2014-08-15 2016-02-17 阿里巴巴集团控股有限公司 一种外部系统稳定性检测方法及设备
EP3298432B1 (en) * 2015-05-18 2019-07-10 Lasermotive, Inc. Multi-layered safety system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101474481A (zh) * 2009-01-12 2009-07-08 北京科技大学 情感机器人系统
CN201654862U (zh) * 2010-01-14 2010-11-24 李元金 校车行驶记录监控仪
US20130196649A1 (en) * 2012-01-27 2013-08-01 Qualcomm Incorporated Mobile device to detect unexpected behaviour
CN104658140A (zh) * 2013-11-21 2015-05-27 青岛唐鹏钢结构工程有限公司 一种应用于防盗门的多路检测智能防盗报警器
CN106019267A (zh) * 2015-03-31 2016-10-12 新明和工业株式会社 运动物体检测装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109143043A (zh) * 2018-09-30 2019-01-04 深圳市盈科互动科技有限公司 控制器io输入防抖检测方法、系统、装置及存储介质
CN109143043B (zh) * 2018-09-30 2021-04-16 深圳市盈科互动科技有限公司 控制器io输入防抖检测方法、系统、装置及存储介质
CN112051621A (zh) * 2019-08-15 2020-12-08 河南紫联物联网技术有限公司 一种房间有人无人状态的判定方法与装置
CN112051621B (zh) * 2019-08-15 2024-03-15 河南紫联物联网技术有限公司 一种房间有人无人状态的判定方法与装置

Also Published As

Publication number Publication date
CN108254801A (zh) 2018-07-06
TW201823932A (zh) 2018-07-01

Similar Documents

Publication Publication Date Title
US12035106B2 (en) Machine learning model capability assessment
Kotus et al. Detection and localization of selected acoustic events in acoustic field for smart surveillance applications
US11047843B2 (en) Atmospheric environment monitoring apparatus detecting failure of atmospheric environment sensor, and method for detecting failure of atmospheric environment sensor
KR20170031985A (ko) 공기조화시스템의 고장 검출 및 진단 방법
RU2015118362A (ru) Способ и система обнаружения занятости
WO2019196205A1 (zh) 外语教学评价信息生成方法以及装置
JP7056823B2 (ja) ローカル分析の監視のシステム及び方法
CN113920673B (zh) 一种室内火情智能监测方法及系统
KR102863001B1 (ko) 가스 누출 조기 경보 방법, 서버 및 시스템
CN104849352A (zh) 一种隔声性能检测系统
CN114255784A (zh) 一种基于声纹识别的变电站设备故障诊断方法及相关装置
CN109443525A (zh) 一种设备异响检测系统及检测方法
US10739313B2 (en) Acoustic profile recognition for discriminating between hazardous emissions and non-hazardous emissions
WO2023093919A2 (zh) 一种洗衣机异音检测方法、装置、电子设备及存储介质
CN111276140A (zh) 语音命令识别方法、装置、系统及存储介质
US20250094540A1 (en) Apparatus, method, and computer program for determining normalcy of data of digital twin model
WO2018121330A1 (zh) 一种检测方法、相关装置和系统
CN115098293A (zh) 物联网终端设备用的自动测试系统
CN111627460A (zh) 环境混响检测方法、装置、设备与计算机可读存储介质
CN106153363B (zh) 一种基于声像监测的机械设备故障自动识别方法
CN111121232B (zh) 一种空调产线检测方法、装置及计算机可读存储介质
CN114999097B (zh) 一种格栅吊顶内感烟火灾探测器有效性的评估方法及系统
CN116859019A (zh) 居民用燃气的安全检测方法、系统、装置及存储介质
US11557188B2 (en) Gunshot detection device, system and method
US20200202865A1 (en) Control method, voice interaction apparatus, voice recognition server, non-transitory storage medium, and control system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17889096

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17889096

Country of ref document: EP

Kind code of ref document: A1