[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

WO2019136733A1 - Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale - Google Patents

Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale Download PDF

Info

Publication number
WO2019136733A1
WO2019136733A1 PCT/CN2018/072547 CN2018072547W WO2019136733A1 WO 2019136733 A1 WO2019136733 A1 WO 2019136733A1 CN 2018072547 W CN2018072547 W CN 2018072547W WO 2019136733 A1 WO2019136733 A1 WO 2019136733A1
Authority
WO
WIPO (PCT)
Prior art keywords
face recognition
recognition device
head model
angle
slide assembly
Prior art date
Application number
PCT/CN2018/072547
Other languages
English (en)
Chinese (zh)
Inventor
蒋美琼
林魁
邓杨鹏
Original Assignee
福建联迪商用设备有限公司
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 福建联迪商用设备有限公司 filed Critical 福建联迪商用设备有限公司
Priority to CN201880000010.1A priority Critical patent/CN108235769B/zh
Priority to PCT/CN2018/072547 priority patent/WO2019136733A1/fr
Publication of WO2019136733A1 publication Critical patent/WO2019136733A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification

Definitions

  • the present invention relates to the field of face recognition testing technology, and in particular, to a performance testing method and a testing device for a face recognition device.
  • the face recognition device is widely used in identity recognition, as a device for actively recognizing a face (that is, a person who is not forcibly recognized to visually look ahead), therefore, it is necessary to perform a recognition accuracy test on each angle of the produced device, and the test is qualified. Put into the market.
  • the performance test of the face recognition device is mostly performed by using a certain camera device to perform a plurality of test angles on a head model with a face, obtaining a corresponding image, and then importing the corresponding image into the test device in batches.
  • the face recognition device to be tested is tested by a software method, and it is determined whether the recognition degree of the imported image by the face recognition device to be tested reaches a certain threshold. If it is reached, the face recognition device to be tested is qualified.
  • This method is highly efficient, it also has the following drawbacks:
  • the image captured by the camera device is processed internally by the camera device, and there is more or less distortion, so it cannot be exactly the same as the real scene, and thus affects the performance test result of the face recognition device;
  • the face recognition device to be tested by software is tested. However, it does not take into account that the hardware of the face recognition device itself may affect the performance test results of the face recognition device, such as lens parameters.
  • the technical problem to be solved by the present invention is to provide a test method and a test apparatus for detecting whether the performance of the face recognition device is qualified.
  • the technical solution adopted by the present invention is:
  • a performance testing method for a face recognition device includes the following steps:
  • the face recognition device to be tested acquires face recognition data when the head model is at a position; the position is composed of a distance and an angle;
  • the similarity between the face recognition data and the standard face recognition data is calculated, and the similarity value of the head model at a position is obtained and recorded.
  • a performance testing device for a face recognition device comprising: a test mechanism consisting of a horizontal movement mechanism and an angle test tool; and a head model disposed on the test mechanism; the head model is set on the face recognition device to be tested
  • the front side also includes a processor coupled to the face recognition device to be tested, the processor being provided with a memory.
  • the invention provides a performance testing method and a testing device for a face recognition device, which acquires face recognition data when a head model is located at a position through a face recognition device to be tested, and a head model with a preset database
  • the corresponding standard face recognition data is used to calculate the similarity, and is used to determine whether the face recognition device to be tested can recognize the head model at the position, and if the face recognition device is tested at multiple positions by repeatedly testing the head model Both can be identified, and if they are all recognized, the performance of the face recognition device to be tested is deemed to be qualified.
  • One of the positions consists of a distance and an angle. The distance of the head model can be adjusted and the angle can be adjusted to simulate several positions, thereby improving the test accuracy.
  • FIG. 1 is a flow chart showing the steps of a performance testing method for a face recognition device according to the present invention
  • FIG. 2 is a schematic structural diagram of a performance testing apparatus for a face recognition device according to the present invention.
  • angle test tooling 21, mounting seat; 22, horizontal rotating device; 221, cylinder; 222, rotating shaft; 23, pitch angle adjusting device; 231, boss; 232, adjusting frame; 2321, horizontal plate; Connecting arm; 24, tilt angle adjusting device;
  • the most critical idea of the present invention is to obtain the face recognition data when the head model is located at a position, and calculate the similarity with the standard face recognition data corresponding to the head model in the preset database, and determine the similarity to be tested. Whether the face recognition device can recognize the head model at the position.
  • a method for testing performance of a face recognition device includes the following steps:
  • the face recognition device to be tested acquires face recognition data when the head model is at a position; the position is composed of a distance and an angle;
  • the similarity between the face recognition data and the standard face recognition data is calculated, and the similarity value of the head model at a position is obtained and recorded.
  • the invention provides a performance testing method for a face recognition device, which acquires face recognition data when a head model is located at a position through a face recognition device to be tested, and a standard corresponding to a head model in a preset database.
  • the face recognition data is used to calculate the similarity, and is used to determine whether the face recognition device to be tested can recognize the head model at the position, and whether the face recognition device can recognize when the head model is in different positions by multiple tests.
  • the performance of the face recognition device to be tested is deemed to be qualified.
  • One of the positions consists of a distance and an angle. The distance of the head model can be adjusted and the angle can be adjusted to simulate several positions, thereby improving the test accuracy.
  • the method further includes the following steps:
  • the location, the face recognition data, and the similarity value are associated.
  • the method further includes the following steps:
  • the face recognition device to be tested respectively obtains face recognition data when the head model is located at each position within the nominal position range.
  • the method further includes the following steps:
  • the angle of the head model is sequentially adjusted.
  • angle of the head model is adjusted in turn, specifically:
  • the two angles of the horizontal angle, the pitch angle, and the tilt angle of the head model remain unchanged, and the angles other than the two angles are sequentially adjusted.
  • the method further includes the following steps:
  • a test report of the face recognition device to be tested is generated.
  • test report of the face recognition device to be tested is used as a test basis for the face recognition device.
  • the method further includes the following steps:
  • the present invention further provides a performance testing apparatus for a face recognition device, comprising: a test mechanism composed of a horizontal movement mechanism 1 and an angle test tool 2; and a head model disposed on the test mechanism;
  • the model is disposed in front of the face recognition device to be tested, and further includes a processor coupled to the face recognition device to be tested, the processor being provided with a memory.
  • the performance testing device for the face recognition device acquires face recognition data when the head model set on the testing mechanism is located at a position by the face recognition device to be tested, and is preset with the memory
  • the standard face recognition data corresponding to the head model in the database is used to calculate the similarity, and is used to determine whether the face recognition device to be tested can recognize the head model at the position, and the test head model is in different positions by multiple tests. Whether the face recognition device can be recognized, if all can be identified, the performance of the face recognition device to be tested is deemed to be qualified.
  • the face recognition device to be tested is configured to acquire face recognition data when the head model is at a position; the position is composed of a distance and an angle.
  • the memory is configured to store standard face recognition data corresponding to the head model.
  • the base further includes a base, and the horizontal moving mechanism 1 and the angle test tool 2 are disposed on the base;
  • the horizontal moving mechanism 1 includes an XY axis translation device 11 and a table 12 slidably disposed on the XY axis translation device, the XY axis translation device 11 is disposed on the base;
  • the angle test tool 2 is fixedly mounted on the worktable 12, and the angle test tool 2 includes a mount 21, a horizontal turning device 22, a pitch angle adjusting device 23, a tilt angle adjusting device 24 and a belt which are sequentially connected from bottom to top.
  • the head model of the face is fixedly connected to the table 12 by the mount 21.
  • the horizontal distance test is performed by the XY-axis translation device, and the horizontal angle, the pitch angle, and the tilt angle are tested by the horizontal rotation device, the pitch angle adjustment device, and the tilt angle adjustment device.
  • the XY axis translation device is slidably coupled to the base.
  • the adjustment range of the XY axis can be increased to accommodate various head models and face recognition devices.
  • the XY axis translation device 11 includes an X-axis slide assembly 111, a motion slider 112, and a Y-axis slide assembly 113; the X-axis slide assembly is slidably mounted on a base, the Y-axis slide The assembly is located on the X-axis slide assembly and is slidably coupled to the X-axis slide assembly by a motion slide that is located on the Y-axis slide assembly and slidably coupled to the Y-axis slide assembly.
  • the X-axis slide rail assembly and the Y-axis slide rail assembly have an upper and lower structural relationship, and are slidably connected by a moving slider.
  • the XY axis translation device includes an X-axis slide assembly, a motion slider and a Y-axis slide assembly; the Y-axis slide assembly is slidably mounted on the base, and the X-axis slide assembly is located on the Y-axis
  • the slide rail assembly is slidably coupled to the Y-axis slide assembly by motion sliding, the table being located on the X-axis slide assembly and slidably coupled to the X-axis slide assembly.
  • the X-axis slide rail assembly and the Y-axis slide rail assembly have an upper and lower structural relationship, and are slidably connected by a moving slider.
  • a motion control board 3 electrically connected to the processor is further included, and the motion control board 3 is electrically connected to the motion slider 112.
  • the motion slider is controlled by the motion control board, thereby moving the X-axis slide assembly or the Y-axis slide assembly located on the motion slider.
  • the motion control board is based on the PC interface. Due to the powerful functions of the PC, the motion controller composed with it has the strongest function.
  • the motion control board inserts the control card into the host of the PC through the PCI slot, and develops using high-level programming languages such as C++, C#, VB, VB.net, and labview; the control provided by the motion control board manufacturer is used in programming.
  • the card API interface function is used to realize the use of the control card resources; the motion control board controls the servo motor or the stepping motor by sending a pulse to control the servo motor or the stepping motor, and by reading the input signal and controlling the output signal, the pair is realized.
  • the X-axis slide assembly is provided with an X-axis sensor connected to the processor; and the Y-axis slide assembly is provided with a Y-axis sensor connected to the processor.
  • the X-axis slide assembly is precisely controlled by providing an X-axis sensor connected to the processor on the X-axis slide assembly and a Y-axis sensor connected to the processor on the Y-axis slide assembly. Or the position of the Y-axis rail assembly.
  • the horizontal rotating device 22 includes a coaxial body 221 and a rotating shaft 222.
  • the rotating shaft 222 is connected to the mounting seat 21 at one end and extends into the cylindrical body 221 at the other end.
  • the rotating shaft 222 drives the cylinder 221 horizontally. 360° rotation.
  • the horizontal 360° rotation is achieved by the above structure, which is suitable for the horizontal 360° test.
  • the pitch angle adjusting device 23 includes a boss 231 and an adjusting frame 232 disposed on the upper surface of the cylinder, the boss 231 is provided with a rotating rod, and the adjusting frame 232 includes a horizontal plate 2321 and two The connecting arms 2322 are respectively disposed on opposite sides of the horizontal plate and extend downward in the vertical direction.
  • the connecting arms are provided with through holes, and the through holes of the two connecting arms are respectively connected to the two ends of the rotating rod.
  • the pitch angle adjustment is realized by the above configuration.
  • the adjustment range of the pitch angle adjusting device is -45° to 45°, wherein the center point of the rotating rod is taken as the origin, and the vertical direction of the rotating rod is 0°.
  • the pitch angle adjustment range is -45° to 45°, which is sufficient for the performance test of most face recognition devices.
  • the boss is integrally formed on the upper surface of the cylinder.
  • the tilt angle adjusting device 24 includes a steering gear disposed on a surface of the horizontal plate and a steering gear control panel connected to the steering gear, and the steering gear control panel is coupled to the processor.
  • the tilt angle adjustment is realized by the above configuration.
  • the distance between the head model and the face recognition device to be tested is 10 cm to 1 m.
  • the distance between the head model and the face recognition device to be tested is 10 cm to 1 m, which is sufficient for the performance test of most face recognition devices.
  • a prompter connected to the processor is further included.
  • the face recognition device cannot recognize the prompt, and the prompting device can make a prompt, and the prompting manner can use an audio prompt, a text prompt, a light prompt, and the like.
  • Embodiment 1 of the present invention is:
  • a method for testing performance of a face recognition device includes the following steps:
  • the face recognition device to be tested acquires face recognition data when the head model is at a position; the position is composed of a distance and an angle;
  • the similarity between the face recognition data and the standard face recognition data is calculated, and the similarity value of the head model at a position is obtained and recorded.
  • the performance testing method of the above face recognition device further includes the following steps:
  • the location, the face recognition data, and the similarity value are associated.
  • the performance testing method of the above face recognition device further includes the following steps:
  • the face recognition device to be tested respectively obtains face recognition data when the head model is located at each position within the nominal position range.
  • the performance testing method of the above face recognition device further includes the following steps:
  • the angle of the head model is sequentially adjusted.
  • the angle of the head model is adjusted in turn, specifically:
  • the two angles of the horizontal angle, the pitch angle, and the tilt angle of the head model remain unchanged, and the angles other than the two angles are sequentially adjusted.
  • the performance testing method of the above face recognition device further includes the following steps:
  • a test report of the face recognition device to be tested is generated.
  • the performance testing method of the above face recognition device further includes the following steps:
  • Embodiment 2 of the present invention is:
  • a performance testing apparatus for a face recognition device includes a test mechanism composed of a horizontal movement mechanism and an angle test tool and a head model disposed on the test mechanism; the head model is set In front of the face recognition device to be tested, a processor connected to the face recognition device to be tested is further included, and the processor is provided with a memory.
  • the memory is configured to store standard face recognition data corresponding to the head model.
  • the face recognition device to be tested is configured to acquire face recognition data when the head model is at a position; the position is composed of a distance and an angle. Through the adjustable distance and angle of the head model, several positions can be simulated to improve the test accuracy.
  • the performance testing device of the face recognition device further includes a base, and the horizontal moving mechanism and the angle test tool are all disposed on the base;
  • the horizontal moving mechanism includes an XY axis translation device and a table slidably disposed on the XY axis translation device, the XY axis translation device being disposed on the base;
  • the angle test tool is fixedly mounted on the workbench, and the angle test tool comprises a mount connected in sequence from the bottom, a horizontal rotation device, a pitch angle adjustment device, a tilt angle adjustment device, and a head model with a human face.
  • the mount is fixedly connected to the workbench.
  • the XY axis translation device is slidably coupled to the base.
  • the XY axis translation device comprises an X-axis slide assembly, a motion slider and a Y-axis slide assembly; the X-axis slide assembly is slidably mounted on the base, and the Y-axis slide assembly is located on the X-axis slide
  • the rail assembly is slidably coupled to the X-axis rail assembly by motion sliding, the table being located on the Y-axis rail assembly and slidably coupled to the Y-axis rail assembly.
  • the performance testing device of the face recognition device described above further includes a motion control board electrically connected to the processor, the motion control board being electrically connected to the motion slider.
  • the X-axis slide assembly is provided with an X-axis sensor connected to the processor; and the Y-axis slide assembly is provided with a Y-axis sensor connected to the processor.
  • the horizontal rotating device comprises a coaxially arranged cylinder and a rotating shaft, the rotating end is connected to the mounting seat, and the other end extends into the cylinder, and the shaft rotates the cylinder to rotate 360° horizontally.
  • the pitch angle adjusting device includes a boss and an adjusting frame disposed on an upper surface of the cylinder, the boss is provided with a rotating rod, and the adjusting frame includes a horizontal plate and two opposite sides respectively disposed on the horizontal plate a connecting arm extending downward in a vertical direction, the connecting arm is provided with a through hole, and the through holes of the two connecting arms are respectively connected to both ends of the rotating rod.
  • the adjustment range of the pitch angle adjusting device is -45° to 45°, wherein the center point of the rotating rod is taken as the origin, and the vertical direction of the rotating rod is 0°.
  • the boss is integrally formed on the upper surface of the cylinder.
  • the tilt angle adjusting device includes a steering gear disposed on a surface of the transverse plate and a steering gear control panel coupled to the steering gear, the steering gear control panel being coupled to the processor.
  • the distance between the head model and the face recognition device to be tested is 10 cm to 1 m.
  • the performance testing device of the face recognition device described above further includes a prompter connected to the processor.
  • test principle of the performance testing device for the face recognition device is:
  • the performance testing device of the face recognition device mainly comprises a three-axis angle system and a two-axis mechanical structure to form a hardware system, and the hardware system communicates with the PC through a serial port to perform testing of various angles and distances.
  • the positioning operation of the device is performed before the hardware system performs the formal test, and the intention is to determine the initialization distance position of the mechanical structure and the initialization angle of the angle system.
  • the PC sends an initial positioning command through the serial port.
  • the motion control board After receiving the command information, the motion control board first locates the front, rear, left and right distances. The specific positioning process is performed by the X-axis sensor and the Y-axis sensor on the mechanical structure to complete the position information. After the distance is completed, the PC sends an angle positioning command. After receiving the angle positioning command information, the motion control board locates the angle system, and the positioning completes the return instruction to the PC. After the PC receives the success, it represents the completion of the entire hardware test system. Positioning. It should be noted that the PC is the processor, and any hardware with processing function can be used instead.
  • the PC sends commands to the motion control board through the serial port.
  • the motion control board controls the motion slider to move to 10cm.
  • the coordinates are (10cm, 0cm), one coordinate is the front and rear coordinates, and one coordinate is the left and right coordinates from the center of the device.
  • the PC controls the face recognition device through the serial port to perform the face recognition operation, that is, the face recognition data when the location is obtained, and returns to the PC, and the PC is from the memory.
  • the face recognition operation that is, the face recognition data when the location is obtained, and returns to the PC, and the PC is from the memory.
  • the motion control board controls the motion slider to move to 1m and multiple points between 10cm-1m.
  • the motion slider controls the motion slider to move to 1m and multiple points between 10cm-1m.
  • Angle test assuming that the nominal face recognition device has a horizontal rotation angle of ⁇ 30°, an up and down pitch angle of ⁇ 20°, and a tilt angle of ⁇ 45°.
  • the horizontal rotation angle is controlled from 0° to 30°.
  • the PC controls the first axis of the three-axis angle to perform a horizontal rotation test of incrementing by 1°. Taking 1° as an example, after rotating to the horizontal 1°, waiting for the PC to receive the recognition result of the face recognition device, the PC receives the recognition result.
  • the angle system continues the incremental angle control test, which is cycled until the test reaches 30°.
  • the PC controls the horizontal angle to return to the 0° state. Then, as described above, the test is similar to a horizontal angle of -30°, an up and down pitch angle of ⁇ 20°, and a tilt angle of ⁇ 45°.
  • the PC After receiving the distance return information, the PC controls the angle and sends a command to the angle system to reach the elevation angle of 1°, for example ( 0°, 1°, 0°), the first coordinate is the horizontal rotation angle, the second coordinate is the pitch angle angle, and the third coordinate is the tilt angle angle.
  • the PC receives the information for the next recognition operation, and controls the face recognition device through the serial port to perform the face recognition operation, and the PC obtains the head model from the pre-established database in the memory.
  • PC software Realize automatic recording. If the returned similarity value is lower than the set threshold, an abnormal prompt sound appears at the same time.
  • the PC after testing the face recognition results at various angles and combined states, the PC automatically outputs a test report through the program to determine the recognition result.
  • the above-mentioned three-axis realizes precise control of angle, and can bear the angle control system of 1Kg-1.5KG; the two-axis mechanical structure cooperates with the three-axis angle control to realize simultaneous control of distance and angle; the program of Labview software can realize communication interaction of serial port The test is completed automatically; the test abnormality will prompt the alarm information; after the test is completed, the test report will be automatically produced.
  • Embodiment 3 of the present invention is:
  • the invention also provides a performance testing device for a face recognition device, comprising: a test mechanism composed of a horizontal movement mechanism and an angle test tool; and a head model disposed on the test mechanism; the head model is set to be tested
  • a processor connected to the face recognition device to be tested is further provided, and the processor is provided with a memory.
  • the memory is configured to store standard face recognition data corresponding to the head model.
  • the face recognition device to be tested is configured to acquire face recognition data when the head model is at a position; the position is composed of a distance and an angle. Through the adjustable distance and angle of the head model, several positions can be simulated to improve the test accuracy.
  • the performance testing device of the face recognition device further includes a base, and the horizontal moving mechanism and the angle test tool are all disposed on the base;
  • the horizontal moving mechanism includes an XY axis translation device and a table slidably disposed on the XY axis translation device, the XY axis translation device being disposed on the base;
  • the angle test tool is fixedly mounted on the workbench, and the angle test tool comprises a mount connected in sequence from the bottom, a horizontal rotation device, a pitch angle adjustment device, a tilt angle adjustment device, and a head model with a human face.
  • the mount is fixedly connected to the workbench.
  • the XY axis translation device is slidably coupled to the base.
  • the XY axis translation device comprises an X-axis slide assembly, a motion slider and a Y-axis slide assembly; the Y-axis slide assembly is slidably mounted on the base, and the X-axis slide assembly is located on the Y-axis slide
  • the rail assembly is slidably coupled to the Y-axis rail assembly by motion sliding, the table being located on the X-axis rail assembly and slidably coupled to the X-axis rail assembly.
  • the performance testing device of the face recognition device described above further includes a motion control board electrically connected to the processor, the motion control board being electrically connected to the motion slider.
  • the X-axis slide assembly is provided with an X-axis sensor connected to the processor; and the Y-axis slide assembly is provided with a Y-axis sensor connected to the processor.
  • the horizontal rotating device comprises a coaxially arranged cylinder and a rotating shaft, the rotating end is connected to the mounting seat, and the other end extends into the cylinder, and the shaft rotates the cylinder to rotate 360° horizontally.
  • the pitch angle adjusting device includes a boss and an adjusting frame disposed on an upper surface of the cylinder, the boss is provided with a rotating rod, and the adjusting frame includes a horizontal plate and two opposite sides respectively disposed on the horizontal plate a connecting arm extending downward in a vertical direction, the connecting arm is provided with a through hole, and the through holes of the two connecting arms are respectively connected to both ends of the rotating rod.
  • the adjustment range of the pitch angle adjusting device is -45° to 45°, wherein the center point of the rotating rod is taken as the origin, and the vertical direction of the rotating rod is 0°.
  • the boss is integrally formed on the upper surface of the cylinder.
  • the tilt angle adjusting device includes a steering gear disposed on a surface of the transverse plate and a steering gear control panel coupled to the steering gear, the steering gear control panel being coupled to the processor.
  • the distance between the head model and the face recognition device to be tested is 10 cm to 1 m.
  • the performance testing device of the face recognition device described above further includes a prompter connected to the processor.
  • the present invention provides a performance testing method and a testing device for a face recognition device, which acquires face recognition data when a head model is located at a position through a face recognition device to be tested, and is configured with a preset database.
  • the standard face recognition data corresponding to the head model is used to calculate the similarity, and is used to determine whether the face recognition device to be tested can recognize the head model at the position, and when the head model is in different positions by multiple tests Whether the face recognition device can be recognized, if all can be identified, the performance of the face recognition device to be tested is deemed to be qualified.
  • One of the positions consists of a distance and an angle. The distance of the head model can be adjusted and the angle can be adjusted to simulate several positions, thereby improving the test accuracy.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

La présente invention concerne le domaine technique des tests de reconnaissance faciale, et en particulier un procédé de test de performance et un appareil de test pour un dispositif de reconnaissance faciale. Le procédé de test de performance pour le dispositif de reconnaissance faciale comprend les étapes suivantes : acquisition par un dispositif de reconnaissance faciale à tester de données de reconnaissance faciale lorsqu'un modèle de tête est à une certaine position, la position étant associée à une certaine distance et un certain angle ; acquisition de données de reconnaissance faciale standard correspondant au modèle de tête dans une base de données prédéfinie ; et calcul de la similarité entre les données de reconnaissance faciale et les données de reconnaissance faciale standard, et obtention et enregistrement de la valeur de similarité relativement au moment où le modèle de tête est à la position en question. L'acquisition des données de reconnaissance faciale lorsque le modèle de tête est situé à une certaine position et le calcul de la similarité avec les données de reconnaissance faciale standard correspondant au modèle de tête dans la base de données prédéfinie permettent de déterminer si le dispositif de reconnaissance faciale à tester peut identifier le modèle de tête à ladite position.
PCT/CN2018/072547 2018-01-15 2018-01-15 Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale WO2019136733A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201880000010.1A CN108235769B (zh) 2018-01-15 2018-01-15 一种人脸识别设备的性能测试方法及测试装置
PCT/CN2018/072547 WO2019136733A1 (fr) 2018-01-15 2018-01-15 Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/072547 WO2019136733A1 (fr) 2018-01-15 2018-01-15 Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale

Publications (1)

Publication Number Publication Date
WO2019136733A1 true WO2019136733A1 (fr) 2019-07-18

Family

ID=62657700

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/072547 WO2019136733A1 (fr) 2018-01-15 2018-01-15 Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale

Country Status (2)

Country Link
CN (1) CN108235769B (fr)
WO (1) WO2019136733A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112329661A (zh) * 2020-11-10 2021-02-05 广东博通科技服务有限公司 一种基于深度卷神经网络的人脸识别装置及使用方法
CN114120382A (zh) * 2020-08-25 2022-03-01 中车株洲电力机车研究所有限公司 一种人脸识别系统的测试方法、装置、电子设备及介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109492523A (zh) * 2018-09-17 2019-03-19 深圳壹账通智能科技有限公司 人脸识别系统性能测试方法、装置、设备和存储介质
CN110059673A (zh) * 2019-05-05 2019-07-26 重庆中科云从科技有限公司 一种人脸识别终端自动化测试系统和方法
CN113392035A (zh) * 2021-08-17 2021-09-14 杭州魔点科技有限公司 一种门禁设备的自动化测试方法和系统
CN115861743B (zh) * 2023-02-20 2023-06-02 上海励驰半导体有限公司 基于车载台架的人脸识别装置测试方法、系统及车机

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102982321A (zh) * 2012-12-05 2013-03-20 深圳Tcl新技术有限公司 人脸库采集方法及装置
CN103020607A (zh) * 2012-12-27 2013-04-03 Tcl集团股份有限公司 一种人脸识别方法及装置
CN106067043A (zh) * 2016-06-01 2016-11-02 重庆中科云丛科技有限公司 一种性能测试方法及系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101615196A (zh) * 2009-07-28 2009-12-30 公安部第一研究所 千万级一对多人脸识别产品的测试系统与测试方法
CN107506702B (zh) * 2017-08-08 2020-09-11 江西高创保安服务技术有限公司 基于多角度的人脸识别模型训练及测试系统及方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102982321A (zh) * 2012-12-05 2013-03-20 深圳Tcl新技术有限公司 人脸库采集方法及装置
CN103020607A (zh) * 2012-12-27 2013-04-03 Tcl集团股份有限公司 一种人脸识别方法及装置
CN106067043A (zh) * 2016-06-01 2016-11-02 重庆中科云丛科技有限公司 一种性能测试方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114120382A (zh) * 2020-08-25 2022-03-01 中车株洲电力机车研究所有限公司 一种人脸识别系统的测试方法、装置、电子设备及介质
CN114120382B (zh) * 2020-08-25 2024-05-03 中车株洲电力机车研究所有限公司 一种人脸识别系统的测试方法、装置、电子设备及介质
CN112329661A (zh) * 2020-11-10 2021-02-05 广东博通科技服务有限公司 一种基于深度卷神经网络的人脸识别装置及使用方法
CN112329661B (zh) * 2020-11-10 2022-06-03 广东博通科技服务有限公司 一种基于深度卷神经网络的人脸识别装置及使用方法

Also Published As

Publication number Publication date
CN108235769A (zh) 2018-06-29
CN108235769B (zh) 2020-01-21

Similar Documents

Publication Publication Date Title
WO2019136733A1 (fr) Procédé de test de performance et appareil de test pour dispositif de reconnaissance faciale
CN111331367B (zh) 智能装配控制系统
CN107891414B (zh) 机器人系统
WO1991013318A1 (fr) Procede et appareil de mesure d'angles de façonnage
US10571237B2 (en) CMM with object location logic
CN106323167B (zh) 一种基于图像识别的智能扫描在线测量系统和测量方法
WO2015120734A1 (fr) Dispositif de test spécial et procédé pour corriger une piste de soudage sur la base d'une vision de machine
JP5135672B2 (ja) レーザ照射状態の検出方法およびレーザ照射状態検出システム
CN109202912A (zh) 一种基于单目深度传感器和机械臂配准目标轮廓点云的方法
WO2016151668A1 (fr) Dispositif d'enseignement et procédé de génération d'informations de commande
CN109991019B (zh) 汽车视野校核方法
CN114515924A (zh) 一种基于焊缝识别的塔脚工件自动焊接系统及方法
JP5001330B2 (ja) 曲面部材計測システム及び方法
JP2003511778A (ja) マイクロポジショニングシステム
CN111571596B (zh) 利用视觉修正冶金接插装配作业机器人误差的方法及系统
CN114833832A (zh) 一种机器人手眼标定方法、装置、设备及可读存储介质
US11992739B1 (en) Apparatus and method for measuring lie and loft of a golf club
CN117006972A (zh) 物体检测方法、装置、设备、三维扫描系统及存储介质
WO2022043979A1 (fr) Programme, procédé et système
JP5145291B2 (ja) レーザービジョンモジュール及びこれを用いた非接触計測装置
CN114571199A (zh) 一种锁螺丝机及螺丝定位方法
CN112540564A (zh) 一种配电变压器的接线系统及接线方法
CN108153188B (zh) 一种基于脚本数据的调整检查流程切换方法
TWI730664B (zh) 雷射打標系統及其控制方法
US20190103125A1 (en) System and method for dynamic optical microphone

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: 18899110

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: 18899110

Country of ref document: EP

Kind code of ref document: A1