A non-intrusive sensing terminal device based on sound and anonymous visual modalities
By integrating multimodal sensing components with non-invasive sensing terminal devices that utilize both sound and anonymous visual modalities for local data processing and low-latency transmission, the privacy protection and response lag issues of campus security equipment are resolved, enabling early identification and timely intervention of bullying behavior.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUAZHONG NORMAL UNIV
- Filing Date
- 2026-03-19
- Publication Date
- 2026-06-19
AI Technical Summary
Existing campus security equipment cannot be deployed on a large scale for facial recognition due to privacy concerns, creating security blind spots. Furthermore, its functions are limited and its response is slow, making it difficult to meet the needs for early intervention in bullying.
It adopts a non-intrusive sensing terminal device based on sound and anonymous visual modalities, integrates sound and anonymous visual multimodal sensing components, and achieves local data processing and low-latency transmission through the collaborative control of the StarSpark K01 and D02 development boards. It collects human skeletal posture rather than facial information, and performs multimodal data fusion recognition in combination with the logic operation module.
It effectively distinguishes between normal playfulness and bullying behavior, improves identification accuracy and response speed, achieves a balance between privacy protection and perception needs, and intervenes in bullying behavior in a timely manner.
Smart Images

Figure CN122245016A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of campus security equipment technology, specifically a non-invasive sensing terminal device based on sound and anonymous visual modalities. Background Technology
[0002] School bullying is a prominent problem that endangers the physical and mental health of teenagers and the safety and stability of schools. It is prevalent in areas that are difficult to cover by traditional security measures, such as restrooms, stairwell corners, and dormitory corridors. Current technology relies on high-definition cameras to collect video data and judges abnormal behavior through manual patrols or simple image recognition algorithms.
[0003] However, due to privacy concerns, it is impossible to deploy video surveillance equipment with facial recognition capabilities on a large scale, creating security blind spots. At the same time, existing campus security sensing equipment often suffers from limitations such as limited functionality, slow response, and insufficient privacy protection, making it difficult to meet the actual needs for early intervention in bullying.
[0004] Therefore, there is a need to provide a non-intrusive sensing terminal device based on sound and anonymous visual modalities. Summary of the Invention
[0005] The purpose of this invention is to provide a non-intrusive sensing terminal device based on sound and anonymous visual modalities to solve the problems mentioned in the background art, such as the inability to deploy video surveillance equipment with face capture function on a large scale due to privacy protection, thus creating security blind spots; at the same time, existing campus security sensing devices have defects such as single function, slow response, and insufficient privacy protection, making it difficult to meet the actual needs of early intervention for bullying behavior.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a non-invasive sensing terminal device based on sound and anonymous visual modalities, comprising a housing and a base plate, wherein a door is hinged to the front of the housing, an edge center component is installed inside the housing, a shell is fixedly installed on the outside of the base plate, a sensing component is fixedly installed on the inner side of the base plate, an installation component is fixedly connected to the outer side of the base plate, and an alarm component is fixedly installed on the top of the shell.
[0007] Preferably, ventilation grilles and wiring holes are provided on both sides of the enclosure, and locking plates are fixedly installed at the four corners of the enclosure. The ventilation grilles and wiring holes on both sides of the enclosure not only ensure the heat dissipation effect of the edge and center components, but also facilitate the neat arrangement of the wiring.
[0008] Preferably, the edge center component includes a first power module and a motherboard, both of which are installed inside the housing. One end of the first power module is electrically connected to a first heat dissipation module, which is installed on the inner wall of the housing. This ensures the continuous and stable operation of the edge center component, solves the problems of unstable power supply and poor heat dissipation of the core processing unit in existing equipment, and provides a reliable guarantee for local data processing.
[0009] Preferably, a logic operation module, a controller, and a signal receiving module are sequentially mounted on the surface of the motherboard. The motherboard is electrically connected to the first power module. The motherboard integrates the logic operation module, the controller, and the signal receiving module to realize the integration of data reception, processing, and instruction issuance, which simplifies the internal structure of the edge center component and improves the efficiency of data processing and instruction transmission.
[0010] Preferably, the mounting assembly includes a mounting bracket, which is fixedly mounted on the outside of the base plate. Each of the four corners of the mounting bracket has two fixing holes, and the mounting bracket is made of angle iron welded together.
[0011] Preferably, the sensing component includes a signal transmission module, a second power supply module, a StarSpark K01 development board, and a StarSpark D02 development board. The StarSpark K01 and StarSpark D02 development boards are electrically connected and installed in the middle of the inner side of the base plate. The second power supply module and the signal transmission module are both fixedly installed on the base plate, and the StarSpark K01 and StarSpark D02 development boards are both electrically connected to the second power supply module.
[0012] Preferably, the StarSpark K01 and StarSpark D02 development boards are electrically connected via wires to a filtering module, millimeter-wave radar, voltage regulator module, signal transmission module, and INMP441 omnidirectional microphone. One side of the INMP441 omnidirectional microphone is electrically connected to a DFRobot high-sensitivity sound recognition module. The filtering module, millimeter-wave radar, voltage regulator module, signal transmission module, INMP441 omnidirectional microphone, and DFRobot high-sensitivity sound recognition module are all fixedly mounted on the base plate. The StarSpark development boards are connected to various sensing sub-modules via wires to achieve comprehensive acquisition of sound and anonymous visual multimodal data. The INMP441 omnidirectional microphone works in conjunction with the DFRobot high-sensitivity sound recognition module to improve sound perception accuracy. All modules are fixed to the base plate to ensure stable installation.
[0013] Preferably, the front of the housing has several heat dissipation and sound vents, and the bottom of the housing has a wire passage.
[0014] Preferably, a voice broadcast module is fixedly installed inside the housing, and a second heat dissipation module is fixedly installed on one side of the connection between the voice broadcast module and the housing. The second heat dissipation module inside the housing dissipates heat from the voice broadcast module to prevent it from overheating and being damaged during long-term operation. The voice broadcast module can actively issue warning voices, which is more effective in intervention than a single sound and light alarm.
[0015] Preferably, the alarm component includes a mounting plate, which is fixedly mounted on the top of the housing. An audible and visual alarm is mounted on the top of the mounting plate. The audible and visual alarm can emit visual and auditory alarm signals simultaneously, forming a dual warning. Compared with a single alarm method, it has a stronger deterrent effect and can quickly attract the attention of the surrounding area or remind security personnel.
[0016] Compared with the prior art, the beneficial effects of the present invention are: integrating sound and anonymous visual multimodal perception components, and through the collaborative control of the Starflash K01 development board and D02 development board, making full use of the complementarity of different modal data, effectively distinguishing normal playfulness from bullying behavior, solving the problems of easy interference and high false alarm rate of single modal perception in the prior art, and significantly improving the accuracy and reliability of bullying behavior recognition; An anonymous visual perception design that only collects human skeletal postures avoids collecting private information such as facial and limb details. This not only makes up for the lack of key behavioral features in pure acoustic vibration perception, but also solves the privacy leakage problem of traditional visual monitoring equipment, thus achieving an organic unity between privacy protection and perception needs. Local data processing is achieved through the edge center component inside the enclosure, eliminating the need for centralized cloud computing. At the same time, low-latency data transmission is achieved by relying on the signal transmission module of the StarFlash development board. The entire chain response from data acquisition to alarm triggering is rapid, solving the problems of high transmission and processing latency and inability to intervene in a timely manner in existing technologies. It can trigger an effective response within the golden time when bullying occurs. Attached Figure Description
[0017] Figure 1 This is a three-dimensional structural diagram of the present invention; Figure 2 This is the front view of the present invention; Figure 3 This is a diagram of the sensing components of the present invention; Figure 4 This is a partial exploded view of the present invention; Figure 5 This is a schematic diagram of the interior of the outer casing of the present invention; Figure 6 This is a schematic diagram of the external casing of the present invention; Figure 7 This is a schematic diagram of the internal structure of the housing of the present invention; Figure 8 This is a top view of the present invention.
[0018] In the diagram: 1. Cabinet; 2. Cabinet door; 3. Ventilation grille; 4. Wiring hole; 5. Locking plate; 6. Mainboard; 7. Logic module; 8. Controller; 9. Signal receiving module; 10. First power supply module; 11. First heat dissipation module; 12. Base plate; 13. Mounting bracket; 14. Fixing hole; 15. StarSpark K01 development board; 16. StarSpark D02 development board; 17. Second power supply module; 18. Filtering module; 19. Millimeter-wave radar; 20. Voltage regulator module; 21. Signal transmission module; 22. INMP441 omnidirectional microphone; 23. DFRobot high-sensitivity sound recognition module; 24. Outer shell; 25. Heat dissipation and sound outlet; 26. Wiring port; 27. Second heat dissipation module; 28. Voice broadcast module; 29. Mounting plate; 30. Audible and visual alarm. Detailed Implementation
[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
[0020] Please see Figure 1-8 The present invention provides a non-invasive sensing terminal device based on sound and anonymous visual modalities, including a housing 1 and a base plate 12. A door 2 is hinged to the front of the housing 1. An edge center component is installed inside the housing 1. A shell 24 is fixedly installed on the outside of the base plate 12. A sensing component is fixedly installed on the inside of the base plate 12. An installation component is fixedly connected to the outside of the base plate 12. An alarm component is fixedly installed on the top of the shell 24.
[0021] Furthermore, ventilation grilles 3 and wiring holes 4 are provided on both sides of the enclosure 1. Locking plates 5 are fixedly installed at the four corners of the enclosure 1. The ventilation grilles 3 on both sides of the enclosure 1 provide heat dissipation channels for the internal edge center components, reducing the operating temperature of components such as the motherboard 6 and the first power module 10 through air circulation. The wiring holes 4 are used to organize power supply lines and signal lines, avoiding messy lines that may affect the operation or maintenance efficiency of the equipment. The locking plates 5 at the four corners of the enclosure 1 are fixed to the mounting carrier with bolts to ensure that the enclosure 1 and the internal edge center components are installed firmly, providing a stable hardware environment for data processing and signal transmission.
[0022] Furthermore, the edge center assembly includes a first power module 10 and a motherboard 6. Both the first power module 10 and the motherboard 6 are installed inside the housing 1. One end of the first power module 10 is electrically connected to a first heat dissipation module 11, which is installed on the inner wall of the housing 1. In the edge center assembly, the first power module 10 provides a stable operating voltage for core components such as the motherboard 6, logic operation module 7, and controller 8. The first heat dissipation module 11 is installed on the inner wall of the housing 1 and forms a heat exchange channel with the ventilation grille 3 to reduce the operating temperature of the first power module 10, avoid overheating that could lead to power outages or module damage, and ensure the continuous and stable operation of the edge center assembly.
[0023] Furthermore, a logic operation module 7, a controller 8, and a signal receiving module 9 are sequentially mounted on the surface of the motherboard 6. The motherboard 6 is electrically connected to the first power module 10. As the core mounting carrier of the edge center component, the motherboard 6 integrates the logic operation module 7, the controller 8, and the signal receiving module 9 sequentially on its surface, and is electrically connected to the first power module 10. During operation, the signal receiving module 9 receives the structured data transmitted by the sensing component and forwards it to the logic operation module 7. The logic operation module 7 executes a multimodal fusion recognition algorithm to determine whether bullying behavior exists. The controller 8 generates an alarm command based on the determination result and sends it to the relevant components. The three work together to complete the task.
[0024] Furthermore, the installation components include a mounting bracket 13, which is fixedly installed on the outside of the base plate 12. Each of the four corners of the mounting bracket 13 has two fixing holes 14. The mounting bracket 13 is made of angle iron and is fixedly installed on the outside of the base plate 12. The fixing holes 14 at its four corners are used to connect with the mounting carrier such as the wall or ceiling by bolts.
[0025] Furthermore, the sensing component includes a signal transmission module 21, a second power supply module 17, a StarSpark K01 development board 15, and a StarSpark D02 development board 16. The StarSpark K01 development board 15 and the StarSpark D02 development board 16 are electrically connected and installed in the middle of the inner side of the base plate 12. The second power supply module 17 and the signal transmission module 21 are both fixedly installed on the base plate 12. The StarSpark K01 development board 15 and the StarSpark D02 development board 16 are both electrically connected to the second power supply module 17. During operation, the StarSpark K01 development board 15 collects the data collected by each sensing sub-module and transmits it to the StarSpark D02 development board 16 via the signal transmission module 21. The data is then sent to the signal receiving module 9 of the edge center component with low latency through the StarSpark wireless communication protocol, thereby realizing efficient transmission and collaborative processing of sensing data.
[0026] Furthermore, the StarSpark K01 development board 15 and StarSpark D02 development board 16 are electrically connected via wires to a filter module 18, a millimeter-wave radar 19, a voltage regulator module 20, a signal transmission module 21, and an INMP441 omnidirectional microphone 22. One side of the INMP441 omnidirectional microphone 22 is electrically connected to a DFRobot high-sensitivity sound recognition module 23. The filter module 18, millimeter-wave radar 19, voltage regulator module 20, signal transmission module 21, INMP441 omnidirectional microphone 22, and DFRobot high-sensitivity sound recognition module 23 are all fixedly mounted on the base plate 12. During operation, the INMP441 omnidirectional microphone 22 works with the DFRobot high-sensitivity sound recognition module 23 to collect sound signals and transmit them to the model for processing. The millimeter-wave radar 19 captures signals and extracts impact features. The voltage regulator module 20 stabilizes the input voltage to ensure power supply to each module. The structured data collected by each sub-module is integrated by the StarSpark development board and transmitted to the edge center component, providing multimodal data support for bullying behavior determination.
[0027] Furthermore, the front of the housing 24 is provided with several heat dissipation and sound outlet holes 25, and the bottom of the housing 24 is provided with a wiring port 26. The heat dissipation and sound outlet holes 25 on the front of the housing 24 provide heat dissipation channels for internal components and work with the second heat dissipation module 27 to achieve air circulation and cooling. The wiring port 26 at the bottom of the housing 24 is used to organize the power supply lines and signal lines of the sensing components, avoid the lines from being tangled or exposed, ensure the safe operation of the equipment and the standardization of wiring, and at the same time provide a channel for the sound transmission of the voice broadcast module 28.
[0028] Furthermore, a voice broadcast module 28 is fixedly installed inside the outer casing 24. A second heat dissipation module 27 is fixedly installed on one side of the connection between the voice broadcast module 28 and the outer casing 24. During operation, the second heat dissipation module 27 monitors the operating temperature of the voice broadcast module 28 in real time and reduces its operating temperature through heat exchange to avoid overheating damage. When an alarm command is received from the edge center component, the voice broadcast module 28 starts and plays a preset warning voice, which is transmitted to the outside through the heat dissipation sound outlet 25 to achieve on-site deterrence and intervention against bullying behavior.
[0029] Furthermore, the alarm component includes a mounting plate 29, which is fixedly mounted on the top of the housing 24. An audible and visual alarm 30 is mounted on the top of the mounting plate 29. The mounting plate 29 provides a stable mounting carrier for the audible and visual alarm 30, ensuring its fixed position. When an alarm command is received from the edge center component, the audible and visual alarm 30 simultaneously emits a high-decibel alarm sound and flashing lights, attracting attention from the surrounding area or alerting security personnel through dual auditory and visual warnings, quickly preventing the escalation of bullying behavior.
[0030] In this embodiment, during use: according to installation requirements, the sensing unit consisting of the base plate 12 and the outer shell 24 is fixed to the bathroom wall and the corridor ceiling respectively through the fixing holes 14 of the mounting bracket 13; the box 1 is fixed to the wall of the corridor weak current room through the locking piece 5, completing the installation of the edge center component; the power supply line and signal line are neatly arranged through the wiring hole 4 of the box 1 and the wiring port 26 of the outer shell 24; the first power module 10 and the second power module 17 are connected to 220V mains power to ensure continuous power supply to the equipment; the voltage regulator module 20 starts working, stabilizing the input voltage to the working voltage adapted to each module; the equipment power is turned on, and the StarShine K01 development board 15 and StarShine D02 development board 16 automatically establish electrical connection after being powered on. Next, the system firmware completes initialization; the controller 8 on the motherboard 6 starts self-test, the logic operation module 7 and the signal receiving module 9 enter standby mode, the first heat dissipation module 11 and the second heat dissipation module 27 start simultaneously, and the staff completes parameter configuration through the equipment's debugging interface, sets the anonymous visual acquisition function of the toilet sensing unit to be turned off, and the anonymous visual acquisition function of the corridor sensing unit to be enabled; sets the sound recognition threshold and vibration impact acceleration threshold, and configures the alarm volume and flashing frequency of the sound and light alarm 30; the INMP441 omnidirectional microphone 22 and the DFRobot high-sensitivity sound recognition module 23 continuously collect sound signals in the toilet and transmit them to the model in real time to complete the decibel value calculation and acoustic feature extraction.When students push and shove each other, the millimeter-wave radar 19 captures the impact vibration signal from the wall and extracts characteristic data such as impact acceleration and vibration frequency. The StarSpark K01 development board 15 receives the structured sound and vibration characteristic data and sends it to the StarSpark D02 development board 16 via the signal transmission module 21. The data is then transmitted with low latency to the signal receiving module 9 inside the enclosure 1 via the StarSpark wireless communication protocol. In addition to the data acquisition work of the INMP441 omnidirectional microphone 22, the DFRobot high-sensitivity sound recognition module 23, and the millimeter-wave radar 19, the anonymous vision module simultaneously collects human skeletal posture data in the corridor, extracting only limb movement data. The motion trajectory features, without collecting private information such as facial features, are transmitted via wire to the StarShine K01 development board 15. The StarShine K01 development board 15 integrates the three types of feature data—sound, vibration, and anonymous visual—and then transmits them to the signal receiving module 9 via the StarShine D02 development board 16, achieving synchronous aggregation of multimodal data. The signal receiving module 9 transmits the received structured data to the logic operation module 7. The logic operation module 7 calls a preset multimodal fusion recognition algorithm to perform collaborative analysis of the data. Combining the decibel value of the sound signal and the impact characteristics of the vibration signal, it determines whether there is physical conflict or verbal bullying, and superimposes the anonymous visual limb... Body posture characteristics are used to further improve the accuracy of judgment and avoid misjudging normal playfulness as bullying. When the conflict in the restroom escalates, the sound signal reaches 90dB and the vibration signal impact acceleration reaches 8m / s². The logic operation module 7 calculates and confirms that the bullying judgment conditions are met, and generates a high-confidence alarm command. After receiving the command, the controller 8 synchronously triggers the local linkage and remote notification process. The controller 8 sends a command to the voice broadcast module 28 inside the shell 24 through the signal transmission module 21. The voice broadcast module 28 starts and plays the warning voice "Please do not engage in conflict. School bullying is unacceptable" through the heat dissipation sound hole 25. At the same time The audible and visual alarm 30 on the top of the casing 24 is activated, emitting a high-decibel alarm sound and flashing red light, creating a dual deterrent effect through hearing and sight. The mainboard 6 transmits the alarm information to the campus security center platform via signal lines. Simultaneously, the logic operation module 7 continuously receives real-time data transmitted by the sensing components, dynamically updating the conflict status information. After receiving the alarm information, security personnel quickly arrive at the scene via the corridor to stop the bullying behavior. After the on-site handling is completed, a handling completion command is sent through the equipment's interface. Upon receiving the command, the controller 8 controls the voice broadcast module 28 and the audible and visual alarm 30 to stop working, and the equipment resumes data acquisition mode.
[0031] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A non-invasive sensing terminal device based on sound and anonymous visual modalities, comprising a housing (1) and a base plate (12), characterized in that: The front of the enclosure (1) is hinged with a door (2), an edge center assembly is installed inside the enclosure (1), an outer shell (24) is fixedly installed on the outside of the base plate (12), a sensing assembly is fixedly installed on the inside of the base plate (12), an installation assembly is fixedly connected to the outside of the base plate (12), and an alarm assembly is fixedly installed on the top of the outer shell (24).
2. The non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: Both sides of the box (1) are provided with ventilation grilles (3) and wire holes (4), and the four corners of the box (1) are fixedly installed with locking pieces (5).
3. The non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: The edge center assembly includes a first power module (10) and a motherboard (6). Both the first power module (10) and the motherboard (6) are installed inside the housing (1). One end of the first power module (10) is electrically connected to a first heat dissipation module (11), which is installed on the inner wall of the housing (1).
4. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 3, characterized in that: The motherboard (6) is sequentially equipped with a logic operation module (7), a controller (8) and a signal receiving module (9), and the motherboard (6) is electrically connected to the first power module (10).
5. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: The mounting assembly includes a mounting bracket (13), which is fixedly mounted on the outside of the base plate (12). Each of the four corners of the mounting bracket (13) has two fixing holes (14). The mounting bracket (13) is made of angle iron welded together.
6. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: The sensing component includes a signal transmission module (21), a second power supply module (17), a StarSpark K01 development board (15), and a StarSpark D02 development board (16). The StarSpark K01 development board (15) and the StarSpark D02 development board (16) are electrically connected. The StarSpark K01 development board (15) and the StarSpark D02 development board (16) are both installed in the middle of the inner side of the base plate (12). The second power supply module (17) and the signal transmission module (21) are both fixedly installed on the base plate (12). The StarSpark K01 development board (15) and the StarSpark D02 development board (16) are both electrically connected to the second power supply module (17).
7. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 6, characterized in that: The StarSpark K01 development board (15) and StarSpark D02 development board (16) are electrically connected via wires to a filter module (18), a millimeter-wave radar (19), a voltage regulator module (20), a signal transmission module (21), and an INMP441 omnidirectional microphone (22). One side of the INMP441 omnidirectional microphone (22) is electrically connected to a DFRobot high-sensitivity sound recognition module (23). The filter module (18), millimeter-wave radar (19), voltage regulator module (20), signal transmission module (21), INMP441 omnidirectional microphone (22), and DFRobot high-sensitivity sound recognition module (23) are all fixedly mounted on the base plate (12).
8. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: The front of the outer casing (24) has several heat dissipation and sound outlet holes (25), and the bottom of the outer casing (24) has a wire passage (26).
9. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: A voice broadcast module (28) is fixedly installed inside the outer shell (24), and a second heat dissipation module (27) is fixedly installed on one side of the connection between the voice broadcast module (28) and the outer shell (24).
10. A non-intrusive sensing terminal device based on sound and anonymous visual modalities according to claim 1, characterized in that: The alarm assembly includes a mounting plate (29) which is fixedly mounted on the top of the housing (24), and an audible and visual alarm (30) is mounted on the top of the mounting plate (29).