Portable laser interception device and audio reconstruction mode

Abstract

The invention discloses a portable laser interception device and an audio reconstruction mode, and relates to the technical field of laser interception devices.The portable laser interception device comprises a machine body and a heat dissipation assembly, a display panel is arranged at one end of the machine body, a micropore exhaust pipe is arranged between U-shaped pipes, and air guide grooves are formed in the two sides of the machine body. Air pre-cooled by the cooling liquid moves along the interiors of the upper cavity, the air guide groove and the lower cavity to absorb heat emitted by the machine body and cool elements in the machine body, so that the temperature adaptability of the device is improved, compared with traditional air cooling, the air can be pre-cooled in advance, the heat dissipation effect is improved, and compared with traditional liquid cooling, the heat dissipation efficiency is improved. No complex cooling pipeline needs to be arranged, meanwhile, direct contact between elements and outside air can be avoided without passing through the interior of the machine body, the problem that heat rises due to excessive heat absorption caused by long-term contact between cooling liquid and the high-heat machine body in the circulation process, and therefore the heat dissipation effect is reduced can be solved, the use amount of the cooling liquid can be greatly reduced, and the weight of the device is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of laser listening devices, and particularly to a portable laser listening device and an audio reconstruction method. Background Technique

[0002] Laser listening technology is an important branch in the field of laser sensing. Its core is to obtain information through the interaction between laser and the measured target (such as vibration, reflection, scattering, etc.) to achieve non-contact and high-precision sound or signal detection.

[0003] When the existing laser listening devices are working, heat will be generated due to the efficient operation of each component. To prevent overheating, cooling is carried out by air cooling or water cooling. Among them, air cooling depends on the external environmental temperature, and too high external environmental temperature will lead to poor heat dissipation effect, while water cooling requires the layout of cooling pipelines and there is a hidden danger of pipeline leakage. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a portable laser listening device and an audio reconstruction method, which solves the problems raised in the above background technique.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A portable laser listening device includes a body and a heat dissipation component. One end of the body is provided with a display panel, and the other end of the body is provided with a lens. The heat dissipation component includes an outer housing arranged outside the body. An upper cavity is arranged above the body inside the outer housing, and a lower cavity is arranged below the body inside the outer housing. A grid is embedded at the bottom of the outer housing. A coolant box is arranged on the top of the grid, and a fan row is arranged at the bottom of the coolant box. Refrigeration sheets are fixed on both sides of the upper surface of the fan row, and a U-shaped tube is arranged between the refrigeration sheets on the upper surface of the fan row. A microporous exhaust pipe is arranged between the U-shaped tubes. An air delivery pipe is arranged at the position between the refrigeration sheets on the upper surface of the coolant box, and an air outlet box is arranged at the top of the air delivery pipe. An air outlet is arranged on the side of the air outlet box, and an air intake is arranged on the side of the coolant box. Air guiding grooves are opened on both sides of the body.

[0006] Further, the size of the refrigeration sheet is adapted to the internal structure size of the coolant box, and the space between the refrigeration sheets inside the coolant box is filled with coolant.

[0007] Further, the microporous exhaust pipe is completely sunk below the liquid level of the coolant, and the U-shaped tube is higher than the liquid level of the coolant.

[0008] Further, the air outlet box is communicated with the inside of the coolant box through the air delivery pipe, and the air outlet box is located in the upper cavity area.

[0009] Further, the lower cavity is communicated with the internal space of the coolant box located outside the refrigerating sheet through an air intake port.

[0010] Further, the lower cavity is communicated with the upper cavity through an air guiding groove.

[0011] Further, a telescopic sight is arranged on the outer side surface of the outer machine case, and the orientation of the telescopic sight is consistent with the orientation of the lens.

[0012] An audio reconstruction method is applied to the above-mentioned portable laser eavesdropping device, and the audio reconstruction method includes the following steps:

[0013] Step 1: After the infrared laser is shaped and expanded in beam, it is projected onto the surface of a target more than 300 meters away. The diameter of the laser beam expands to 30 cm at the target, covering a sufficient vibration-sensitive area. The speckle field formed by the reflection on the target surface is captured by the receiving optical system, so as to form a high-contrast speckle image.

[0014] Step 2: The high-speed short-wave infrared camera continuously acquires a sequence of speckle images at a frame rate of 1000 fps, and each frame of image contains one million speckle points.

[0015] Step 3: Each frame of image is converted to the frequency domain through two-dimensional Fourier transform to analyze the overall displacement of the speckles, and then the cross-correlation algorithm with sub-pixel accuracy is adopted in the spatial domain to track the microscopic motion of specific speckle clusters.

[0016] Step 4: The speckle displacement time series is converted into a surface vibration signal, and then the pure sound signal is extracted from the mixed vibration mode through cepstrum analysis and blind source separation technology, and finally clear and intelligible voice content is output.

[0017] Further, in the above Step 2, the high-speed short-wave infrared camera is equipped with a narrow-band filter to suppress the interference of background light.

[0018] Further, in the above Step 3, the two methods can detect the speckle displacement at the level of 0.1 pixel, corresponding to the vibration amplitude of 0.3 nm on the target surface.

[0019] The present invention provides a portable laser eavesdropping device and an audio reconstruction method, which have the following beneficial effects:

[0020] 1. For the portable laser eavesdropping device and audio reconstruction method, the air pre-cooled by the coolant moves along the upper cavity, air guide groove, and inside the lower cavity to absorb the heat dissipated by the machine body, thereby cooling the components inside the machine body, improving the temperature adaptability of the device. This method can pre-cool the air in advance compared with traditional air cooling, thus improving the heat dissipation effect. Compared with traditional liquid cooling, it does not require complex cooling pipelines to be laid, and at the same time, it does not pass through the inside of the machine body, avoiding direct contact between components and external air, and can solve the problem that the coolant absorbs excessive heat during the long-term contact with the high-temperature machine body during the coolant circulation process, resulting in an increase in heat and a decrease in the heat dissipation effect. It can also greatly reduce the coolant consumption, thereby reducing the weight of the device.

[0021] 2. For the portable laser eavesdropping device and audio reconstruction method, it breaks through the detection distance of traditional laser eavesdropping equipment. Compared with traditional laser eavesdropping equipment with a detection distance of only about 100 meters, it can perform laser eavesdropping on ultra-long distances of 400 meters and can improve environmental adaptability and signal quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the external shell appearance structure of the present invention;

[0023] Figure 2 It is a schematic diagram of the internal structure of the external shell of the present invention;

[0024] Figure 3 It is a schematic diagram of the bottom structure of the external shell of the present invention;

[0025] Figure 4 It is a schematic diagram of the internal structure of the coolant box of the present invention;

[0026] Figure 5 It is a schematic diagram of the bottom structure of the coolant box of the present invention;

[0027] Figure 6 It is a schematic diagram of the performance index comparison of the present invention.

[0028] In the figure: 1, machine body; 2, display panel; 3, lens; 4, heat dissipation component; 401, external shell; 402, upper cavity; 403, lower cavity; 404, grid; 405, coolant box; 406, fan exhaust; 407, thermoelectric cooler; 408, U-shaped tube; 409, microporous exhaust pipe; 410, air delivery pipe; 411, air outlet box; 5, air guide groove; 6, aiming telescopic sight. SPECIFIC EMBODIMENTS

[0029] The following further describes in detail the embodiments of the present invention with reference to the drawings and examples. The following examples are used to illustrate the present invention, but cannot be used to limit the scope of the present invention.

[0030] As Figures 1 - 6As shown in the figure, the present invention provides a technical solution: a portable laser eavesdropping device, which includes a body 1 and a heat dissipation component 4. One end of the body 1 is provided with a display panel 2, and the other end of the body 1 is provided with a lens 3. The heat dissipation component 4 includes an outer housing 401 disposed outside the body 1. An upper cavity 402 is provided above the body 1 inside the outer housing 401, and a lower cavity 403 is provided below the body 1 inside the outer housing 401. A grid 404 is embedded at the bottom of the outer housing 401. A coolant box 405 is provided at the top of the grid 404, and a fan row 406 is provided at the bottom of the coolant box 405. Refrigeration sheets 407 are fixed on both sides of the upper surface of the fan row 406, and a U-shaped tube 408 is provided between the refrigeration sheets 407 on the upper surface of the fan row 406. A microporous exhaust pipe 409 is provided between the U-shaped tubes 408. An air delivery pipe 410 is provided at the position between the refrigeration sheets 407 on the upper surface of the coolant box 405, and an air outlet box 411 is provided at the top of the air delivery pipe 410. An air outlet is provided on the side of the air outlet box 411, and an air intake is provided on the side of the coolant box 405. Air guiding grooves 5 are provided on both sides of the body 1. The size of the refrigeration sheet 407 is adapted to the internal structure size of the coolant box 405, and the coolant box 405 is filled with coolant between the refrigeration sheets 407. The microporous exhaust pipe 409 is completely submerged below the liquid level of the coolant, and the U-shaped tube 408 is higher than the liquid level of the coolant. The air outlet box 411 is connected to the inside of the coolant box 405 through the air delivery pipe 410, and the air outlet box 411 is located in the area of the upper cavity 402. The lower cavity 403 is connected to the internal space of the coolant box 405 outside the refrigeration sheet 407 through the air intake, and the lower cavity 403 is connected to the upper cavity 402 through the air guiding groove 5. A aiming telescopic sight 6 is provided on the outer side of the outer housing 401, and the orientation of the aiming telescopic sight 6 is the same as that of the lens 3;

[0031] The specific operation is as follows. A handle is provided at the top of the outer housing 401 for lifting and carrying. The outer housing 401 can be installed on a tripod with an angle adjustment function. The distant target is magnified through the aiming telescopic sight 6 for aiming at the detection target. When the device performs laser eavesdropping, heat is generated inside the body 1 due to the operation of various components, and the heat is dissipated into the upper cavity 402 and the lower cavity 403 inside the outer housing 401;

[0032] At this time, the fan row 406 operates to guide the outside air to pass through the middle of the grid 404 and enter the U-shaped tube 408. The air enters the microporous exhaust pipe 409 along the U-shaped tube 408 and is discharged from the micropores on its surface and enters the coolant. The coolant is used to pre-cool the air. The pre-cooled air enters the air outlet box 411 along the air delivery pipe 410 and is discharged into the upper cavity 402 from the air outlet. The cold air moves along the inside of the upper cavity 402 and moves downward from the air guiding grooves 5 on both sides of the body 1 into the lower cavity 403. After the cold air enters the lower cavity 403 from the upper cavity 402, it then passes through the outer side of the refrigeration sheet 407 along the air intake and is discharged from both sides of the grid 404;

[0033] Wherein, the coolant only pre-cools the external air, and the temperature of the external air is much lower than the heat dissipated by the body 1. Therefore, excessive heat absorption of the coolant is avoided, and the coolant can be continuously cooled by the thermoelectric cooler 407, so as to maintain the pre-cooling effect of the coolant on the external air. The heat on the outer side of the thermoelectric cooler 407 is discharged together with the air after heat absorption, thus avoiding the accumulation of heat inside the outer casing 401;

[0034] Wherein, the grille 404 is in the shape of a louver curtain, and its blades are distributed in a three-section manner. Each section of the blade can be manually pushed to adjust the orientation, so that the discharged hot air and the external air guided into the interior of the outer casing 401 can be separated, avoiding the problem that the hot air is discharged and then guided back into the interior of the outer casing 401;

[0035] Based on the above description, the air pre-cooled by the coolant of the present invention moves along the upper cavity 402, the air guide groove 5, and the lower cavity 403 to absorb the heat dissipated by the body 1, thereby cooling the components inside the body 1, improving the temperature adaptability of the device. This method can pre-cool the air in advance compared with the traditional air cooling, thus improving the heat dissipation effect. Compared with the traditional liquid cooling, it does not require a complex cooling pipeline layout, and at the same time, it does not pass through the inside of the body 1, avoiding direct contact between the components and the external air, and can solve the problem that the coolant absorbs excessive heat during the long-term contact with the high-temperature body 1 during the coolant circulation process, resulting in an increase in heat and a decrease in the heat dissipation effect. It can also greatly reduce the coolant consumption, thereby reducing the weight of the device.

[0036] An audio reconstruction method, which is applied to the above-mentioned portable laser eavesdropping device, and the audio reconstruction method includes the following steps:

[0037] Step 1: After being shaped and expanded, the infrared laser is projected onto the surface of a target more than 300 meters away. The diameter of the laser beam expands to 30 cm at the target, covering a sufficient vibration-sensitive area. The speckle field formed by the reflection on the target surface is captured by the receiving optical system, thereby forming a high-contrast speckle image;

[0038] Step 2: The high-speed short-wave infrared camera continuously acquires a sequence of speckle images at a frame rate of 1000 fps. Each frame of image contains about one million speckles. The high-speed short-wave infrared camera is equipped with a narrow-band filter (1532 ± 5 nm) to suppress background light interference, ensuring that high-quality speckle images can be obtained even under daylight conditions;

[0039] Step 3: Each frame of image is transformed into the frequency domain through two-dimensional Fourier transform to analyze the overall displacement of the speckles. Then, in the spatial domain, a cross-correlation algorithm with sub-pixel accuracy is used to track the microscopic motion of a specific speckle cluster. The fusion of the two methods not only ensures the processing speed but also improves the displacement measurement accuracy, and can detect speckle displacements at the 0.1-pixel level, corresponding to a vibration amplitude of about 0.3 nm on the target surface;

[0040] Step 4: Convert the speckle displacement time series into surface vibration signals, and then extract pure acoustic signals from the mixed vibration modes through cepstrum analysis and blind source separation techniques, and finally output clear and intelligible speech content;

[0041] Specifically, the main challenge faced by 400-meter long-distance laser transmission is the atmospheric turbulence effect, which can cause laser beam drift, expansion, and intensity fluctuations, seriously reducing the quality of speckle images. To overcome the above problems, the following methods are adopted:

[0042] Adaptive optical wavefront correction: Integrate a deformable mirror at the transmitting end to correct the wavefront distortion caused by atmospheric disturbances in real time. The system inversely calculates the instantaneous turbulence intensity by analyzing the statistical characteristics of the echo speckles (such as Strehl ratio and speckle contrast), and uses this to drive the deformable mirror for conjugate compensation. Tests show that this technology can stabilize the spot size within 1.5 times the diffraction limit at a distance of 300 meters;

[0043] Dual-band detection-assisted correction: The system simultaneously collects the speckle fields of the 1532nm detection laser and the 1064nm auxiliary laser, and uses the difference in the transmission of the two wavelengths in the atmosphere to construct a turbulence profile model. This method does not require an additional wavefront sensor and can achieve high-precision turbulence compensation only through speckle analysis;

[0044] Polarization diversity reception technology: There are differences in the scattering characteristics of atmospheric particles for light with different polarization states. The system adopts a polarization diversity reception scheme to simultaneously collect speckle images in orthogonal polarization directions, and suppresses polarization-related noise through an optimal combination algorithm to improve signal stability;

[0045] At the same time, in order to cope with the detection reliability in complex environments, the advantages of the local extreme watershed algorithm, the spot dynamic positioning and tracking algorithm, the spot feature extraction and recognition algorithm, and the spot matching and recognition algorithm are organically combined. The specific combination strategy is as follows:

[0046] Local extreme watershed algorithm: After grayscale processing the image, use the method of gradually increasing the threshold to extract features and effectively identify weak speckle displacements;

[0047] Spot dynamic positioning and tracking algorithm: Detect the moving area of the speckle field through background frame difference segmentation technology, which is especially suitable for rapidly changing vibration modes;

[0048] Spot feature extraction and recognition algorithm: Analyze the morphological features of the speckles (such as area, perimeter, circularity, etc.) and screen out the most representative vibration information;

[0049] Spot matching and recognition algorithm: Adopt the normalized correlation coefficient matching method to compare the current speckle pattern with the typical patterns in the database to improve the recognition specificity;

[0050] The system processes the output results of these four algorithms in real time through a quantum computing acceleration module, and uses the D-S evidence theory for decision-level fusion. The final recognition accuracy can reach 98.7%, which is much higher than the performance of a single algorithm;

[0051] Regarding the problems of voice signal attenuation and distortion caused by long-distance detection, a U-Net form of encoding and decoding structure is adopted in the network architecture, with an attention mechanism and an adversarial training strategy embedded in the middle. It can effectively compensate for high-frequency losses and environmental noise, and a multi-sensor information fusion strategy is introduced. When there are multiple scattering points available for detecting the same target, the system will comprehensively analyze the vibration signals of each point, extract the common-mode components (i.e., the target sound signal) through blind source separation technology, and suppress local interference. This method is particularly suitable for dealing with complex vibration scenarios, such as the situation where multiple people are speaking simultaneously or there is environmental mechanical vibration;

[0052] Performance indicators and test results:

[0053] This laser eavesdropping system has undergone strict laboratory tests and field verifications, and all performance indicators have reached or exceeded the design goals. Especially in the aspect of ultra-long-distance detection of >200 meters, it shows excellent stability and reliability. The system performance is elaborated in detail from three dimensions: technical parameters, environmental adaptability, and comparative advantages;

[0054] Key technical indicators:

[0055] The system shows the following key technical indicators under standard test conditions (visibility > 10 km, relative humidity 40% - 60%, no precipitation):

[0056] Detection distance performance: The effective detection distance of the system reaches 200 meters, and the maximum can be extended to 400 meters (the transmitted power needs to be increased). At a distance of 300 meters, the system can stably detect vibrations with an amplitude as low as 0.3 nanometers on the glass surface, and the corresponding minimum detectable sound pressure level is 35 dB SPL (equivalent to the ambient sound level in a quiet indoor environment). In the test, a standard sound source (94 dB SPL @ 1 kHz) is used to generate sound at the target position, and the signal-to-noise ratio of the reconstructed speech by the system reaches 18 dB, and the word recognition rate exceeds 90%;

[0057] Spatial resolution: The spot diameter of the system at a distance of 200 meters is about 30 cm, and the minimum distinguishable vibration area is 5 cm × 5 cm. This characteristic enables the system to selectively monitor specific areas of the window (such as the center of the glass), avoiding low-sensitivity areas such as the frame;

[0058] Frequency response: The effective detection frequency band of the system is 50 Hz - 5 kHz, covering the main frequency range of human speech. Through a special speckle processing algorithm, the system has an enhanced detection ability for low-frequency vibrations (such as below 100 Hz);

[0059] Real-time performance: The end-to-end latency from speckle acquisition to voice output is less than 50 milliseconds, meeting the requirements of real-time monitoring. The system supports simultaneous tracking of 8 independent targets, with dedicated processing threads assigned to each target, ensuring real-time performance under multitasking through intelligent resource scheduling;

[0060] Power consumption and stability: The overall power consumption of the system is controlled within 25W, and it can operate continuously for 72 hours without frequency reduction. Redundant design and intelligent temperature control are adopted for key components (such as lasers and detectors) to ensure long-term working stability;

[0061] Environmental adaptability test:

[0062] Atmospheric condition test: In hazy weather (visibility about 1km), the system can still maintain a detection distance of 250 meters, with the performance only decreasing by 17%, far superior to traditional laser eavesdropping devices (usually decreasing by more than 50%);

[0063] Light interference test: The system can still operate normally under direct sunlight at noon (illuminance > 100,000 lux), and the sunlight interference suppression ratio reaches 60 dB;

[0064] Vibration and shock test: Vibration (5 - 500 Hz, 5 Grms) and mechanical shock (30 G, 11 ms) tests are carried out according to the MIL-STD-810G standard, and the system performance shows no significant degradation;

[0065] Temperature adaptability test: Within the environmental temperature range of 5℃ to +40℃, the performance fluctuation is less than 10%;

[0066] Complex surface test: In addition to standard glass windows, the system can also effectively detect vibrations on the surfaces of various common building materials, including: granite (sensitivity decreases by about 35%), metal plates (decreases by about 25%), wood materials (decreases by about 15%), etc. For different materials, the system has multiple built-in preprocessing algorithms to automatically optimize parameters;

[0067] Analysis of the comparative advantages of this laser eavesdropping system and similar systems:

[0068] Detection distance: The maximum detection distance of commercial laser eavesdropping devices usually does not exceed 100 meters, while this system reaches 400 meters, which is the farthest among known similar technologies. Compared with the speckle tracking technology reported by Bianchi et al., the detection distance is increased by about 6 times;

[0069] Concealment: The system uses 1532nm eye-safe laser, which is completely invisible, and the transmitted power distribution is uniform, without obvious "hot spot" effect, greatly reducing the risk of anti-reconnaissance;

[0070] Multi-target capability: Through beam segmentation technology, the system can simultaneously monitor vibrations in different areas of the same window and distinguish different sound sources (such as multiple people talking), while traditional systems can usually only handle a single vibration source;

[0071] The modules, chips, cameras, etc. used in the above system are integrated and arranged inside the body 1;

[0072] Based on the above description, this audio reconstruction method combined with the portable laser listening device breaks through the detection distance of traditional laser listening equipment. Compared with the traditional laser listening equipment with a listening distance of only about 100 meters, it can perform laser listening at an ultra-long distance of 400 meters, and can improve environmental adaptability and signal quality.

[0073] In summary, when the portable laser listening device and audio reconstruction method are used, first, a handle is set on the top of the housing 401 for carrying, and the outer housing 401 can be installed on a tripod with an angle adjustment function, and a distant target is magnified by the aiming lens 6 to aim at the detection target. When the device performs laser listening, heat is generated inside the body 1 due to the operation of various components, and the heat is dissipated to the upper cavity 402 and the lower cavity 403 inside the outer housing 401;

[0074] At this time, the fan row 406 operates to guide the outside air to pass through the middle of the grid 404 and enter the U-shaped tube 408. The air enters the microporous exhaust pipe 409 along the U-shaped tube 408 and is discharged from the micropores on its surface and enters the coolant. The coolant is used to pre-cool the air. The pre-cooled air enters the air outlet box 411 along the air supply pipe 410 and is discharged from the exhaust port to the interior of the upper cavity 402. The cold air moves along the interior of the upper cavity 402 and moves downward from the air guide grooves 5 on both sides of the body 1 to the lower cavity 403. After the cold air enters the lower cavity 403 from the upper cavity 402, it passes through the outer side of the refrigeration plate 407 along the air receiving port and is discharged from both sides of the grid 404 to take away the heat, thereby achieving heat dissipation.

[0075] The embodiments of the present invention are given for the purpose of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A portable laser eavesdropping device, comprising a body (1) and a heat dissipation component (4), characterized in that: One end of the body (1) is provided with a display panel (2), and the other end of the body (1) is provided with a lens (3). The heat dissipation component (4) includes an outer housing (401) arranged outside the body (1), and an upper cavity (402) is arranged above the body (1) inside the outer housing (401). A lower cavity (403) is arranged below the body (1) inside the outer housing (401), and a grid (404) is embedded at the bottom of the outer housing (401). A coolant box (405) is arranged at the top of the grid (404), and a fan row (406) is arranged at the bottom of the coolant box (405). Refrigerating sheets (407) are fixed on both sides of the upper surface of the fan row (406), and a U-shaped tube (408) is arranged between the refrigerating sheets (407) on the upper surface of the fan row (406). A microporous exhaust pipe (409) is arranged between the U-shaped tubes (408). An air delivery pipe (410) is arranged at the position between the refrigerating sheets (407) on the upper surface of the coolant box (405), and an air outlet box (411) is arranged at the top of the air delivery pipe (410). An air outlet is arranged on the side of the air outlet box (411), and an air intake is arranged on the side of the coolant box (405). Air guiding grooves (5) are formed on both sides of the body (1).

2. The portable laser eavesdropping device according to claim 1, characterized in that: The size of the refrigerating sheet (407) is adapted to the internal structure size of the coolant box (405), and coolant is filled between the refrigerating sheets (407) inside the coolant box (405).

3. The portable laser eavesdropping device according to claim 2, wherein: The microporous exhaust pipe (409) is completely sunk below the liquid level of the coolant, and the U-shaped tube (408) is higher than the liquid level of the coolant.

4. A portable laser eavesdropping device according to claim 1, characterized in that: The air outlet box (411) is communicated with the inside of the coolant box (405) through the air delivery pipe (410), and the air outlet box (411) is located in the area of the upper cavity (402).

5. A portable laser eavesdropping device according to claim 1, characterized in that: The lower cavity (403) is communicated with the internal space of the coolant box (405) located outside the refrigerating sheet (407) through the air intake.

6. A portable laser eavesdropping device according to claim 1, characterized in that: The lower cavity (403) is communicated with the upper cavity (402) through the air guiding groove (5).

7. A portable laser eavesdropping device according to claim 1, characterized in that: A telescopic sight (6) is arranged on the outer side surface of the outer housing (401), and the orientation of the telescopic sight (6) is the same as that of the lens (3).

8. An audio reconstruction method, which is applied to a portable laser eavesdropping device described in any one of claims 1-7, characterized in that: The audio reconstruction method includes the following steps: Step 1: After the infrared laser is shaped and expanded in beam, it is projected onto the surface of a target more than 300 meters away. The diameter of the laser beam expands to 30 cm at the target, covering a sufficient vibration-sensitive area. The speckle field formed by the reflection of the target surface is captured by the receiving optical system, thereby forming a high-contrast speckle image; Step 2: The high-speed short-wave infrared camera continuously acquires a sequence of speckle images at a frame rate of 1000 fps, and each frame of image contains one million speckle points; Step 3: Each frame of image is converted to the frequency domain through two-dimensional Fourier transform to analyze the overall displacement of the speckles, and then the cross-correlation algorithm with sub-pixel accuracy is adopted in the spatial domain to track the microscopic motion of specific speckle clusters; Step 4: The speckle displacement time series is converted into a surface vibration signal, and then the pure sound signal is extracted from the mixed vibration mode through cepstrum analysis and blind source separation technology, and finally clear and intelligible voice content is output.

9. An audio reconstruction method according to claim 8, characterized in that: In the second step, the high-speed short-wave infrared camera is equipped with a narrow-band filter to suppress background light interference.

10. An audio reconstruction method according to claim 8, characterized in that: In the third step, the two methods can detect speckle displacements at the 0.1 pixel level, corresponding to a vibration amplitude of 0.3 nm on the target surface.

Images