Image secure transmission method and system for intelligent glasses

Abstract

The invention belongs to the technical field of image secure transmission, and particularly relates to an image secure transmission method and system for intelligent glasses, and the method comprises the steps: synchronously collecting video stream data, motion posture data and system state data through a multi-source sensor, and carrying out the spatial-temporal feature alignment; calculating a perception disturbance coupling factor corresponding to each frame of image according to the motion attitude data, the current remaining power percentage and the channel received signal strength indication; calculating an encryption mask threshold value of each frame of image based on the perceptual disturbance coupling factor and the current remaining power percentage; and selectively encrypting the frequency domain coefficient of each frame of image by using the encryption mask threshold and the chaos key stream. According to the method, a dynamic mapping mechanism of the perceptual disturbance coupling factor and the encryption mask threshold value is constructed, so that self-adaptive adjustment of the encryption strength on the motion state of the wearer and the resource state of the equipment is realized, and the endurance time of the equipment is remarkably prolonged while the core privacy security is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image security transmission. More particularly, the present application relates to an image security transmission method and system for smart glasses. BACKGROUND

[0002] With the rapid development of wearable technology, smart glasses, as an important carrier of augmented reality (AR) and first-person view recording, have been widely used in security inspection, remote medical treatment and personal life recording fields. In these application scenarios, the video stream collected by smart glasses in real time often contains sensitive private information such as faces and environmental positions, so the data must be encrypted to prevent leakage.

[0003] However, the existing video encryption technology mainly has the following limitations:

[0004] (1) The contradiction between high energy consumption of full-disk encryption and device endurance: traditional high-intensity encryption algorithms such as AES have high computational complexity, while smart glasses are usually small in size, limited in battery capacity and narrow in heat dissipation space. If high-intensity encryption is performed on each frame of image in full-band, it is easy to cause device computational overload, serious heating and significant reduction of endurance time, which cannot meet the needs of long-term wearing.

[0005] (2) Fixed strategy cannot adapt to dynamic scenes: existing lightweight encryption schemes usually use fixed encryption ratio or pre-set regions of interest, which cannot be dynamically adjusted according to the real-time motion state of the wearer (such as standing, walking, and running) and the current device state (such as power and network signal strength). For example, when running vigorously, the picture shakes, causing the information entropy to increase dramatically. At this time, if the high proportion of encryption is still maintained, it will cause unnecessary waste of computing power. When the power is very low, there is no effective degradation protection mechanism, which may cause the device to shut down prematurely and lose critical data.

[0006] Therefore, there is an urgent need for an adaptive control method to dynamically adjust the encryption strength to achieve the optimal balance between protecting core privacy security and prolonging device endurance. SUMMARY

[0007] To solve the technical problem that the existing video encryption technology cannot achieve the optimal balance between protecting core privacy security and prolonging device endurance, the present application provides solutions in the following aspects.

[0008] In a first aspect, the present invention provides a method for secure image transmission for smart glasses, comprising: synchronously acquiring video stream data, motion posture data, and system status data through multi-source sensors integrated in the smart glasses, and performing spatiotemporal feature alignment on the acquired data; calculating a perceptual perturbation coupling factor corresponding to each frame of image based on the motion posture data, the current remaining battery percentage in the system status data, and the channel received signal strength indication; calculating an encryption mask threshold for each frame of image based on the perceptual perturbation coupling factor and the current remaining battery percentage; and selectively encrypting the frequency domain coefficients of each frame of image using the encryption mask threshold and a preset chaotic key stream to obtain an encrypted video frame.

[0009] Preferably, the simultaneous acquisition of video stream data, motion posture data, and system status data via multi-source sensors integrated in the smart glasses includes: acquiring real-time video streams at a preset frame rate using the front-facing camera of the smart glasses, converting each frame of image data from the RGB color space to the YCrCb color space, and extracting the luminance component Y; acquiring three-axis acceleration and three-axis angular velocity using the inertial measurement unit built into the smart glasses; reading the current remaining battery percentage using the power management chip of the smart glasses, and reading the channel received signal strength indication using the wireless communication module.

[0010] This invention targets the human eye's high sensitivity to changes in brightness by extracting and processing the brightness component. This allows for the achievement of maximum visual obfuscation with minimal computational cost in subsequent encryption steps. Simultaneously, it comprehensively collects motion and system state data, providing data support for multi-dimensional contextual awareness.

[0011] Preferably, the formula for calculating the perceptual perturbation coupling factor corresponding to each frame of image is: In the formula: For the first The perceptual perturbation coupling factor corresponding to the frame image; This is the preset scaling factor; It is the natural logarithm function; For axial index, the value is... , , ; For the first The first frame image corresponding to The absolute value of the axial angular velocity; It is a natural constant; For the first The first frame image corresponding to The rate of change of axial acceleration; For the first The current remaining battery percentage at the time of frame image acquisition; For the first Channel received signal strength indication when frame image is collected.

[0012] The application constructs a perceived disturbance coupling factor, introduces energy and channel state as adjustment levers, when the electric quantity decreases or the signal deteriorates, the denominator decreases, which leads to a significant increase in the value of the perceived disturbance coupling factor, thereby amplifying the influence weight of external motion disturbance on the system, driving the system to make more sensitive strategy adjustment in the low resource and high risk scene, avoiding the rigid fixed encryption strategy.

[0013] Preferably, the first Rate of change of axis acceleration Equal to the first The first The difference between the axis acceleration corresponding to the first The first The difference between the axis acceleration corresponding to the first

[0014] Preferably, the encryption mask threshold is calculated by the following formula: ; In the formula, The encryption mask threshold of the first frame image; Indicates the floor operation; The reference coefficient; The current percentage of residual electric quantity when the first frame image is collected; The hyperbolic tangent function; The perceived disturbance coupling factor corresponding to the first frame image; Indicates the maximum value.

[0015] The application uses the threshold calculation formula to construct a dynamic balance mechanism, wherein the electric quantity term is the main control, which automatically shrinks the encryption range to save energy when the electric quantity is low; the perceived disturbance term is adjusted when detecting violent motion or environmental disturbance to expand the encryption range; this mechanism prevents both the depletion of electric quantity due to excessive protection and the leakage of key private information due to violent shaking of the picture.

[0016] Preferably, the selective encryption of the frequency domain coefficients of each frame image using the encryption mask threshold and the preset chaotic key stream comprises: generating a chaotic key stream with the same length as the image data using the preset chaotic system; performing block discrete cosine transform on the luminance component of the first frame image to obtain a plurality of frequency domain coefficient matrices; for each frequency domain coefficient matrix, traversing the frequency domain coefficients in Zigzag scanning order, and selectively encrypting the frequency domain coefficients of each frame image.

[0017] Preferably, the selective encryption of the frequency domain coefficients of each image frame comprises: for the frequency domain coefficients determined as key privacy coefficients, performing XOR operation and sign bit flipping operation using the chaotic key stream; for the frequency domain coefficients determined as non-key high frequency details, keeping the values unchanged; and performing inverse discrete cosine transformation on the processed frequency domain coefficients to generate the encrypted video frame.

[0018] Preferably, the determination method of the frequency domain coefficients comprises: extracting the scan order of the current frequency domain coefficient; if the scan order is less than the encryption mask threshold, determining the frequency domain coefficient as a key privacy coefficient; and if the scan order is greater than or equal to the encryption mask threshold, determining the frequency domain coefficient as a non-key high frequency detail.

[0019] The present application combines the energy concentration characteristics after discrete cosine transformation with a dynamic threshold, and only processes the low frequency coefficients containing main image information; the value size is confused by XOR operation, and the directionality is destroyed by sign bit flipping, so that the visual structure of the image can be completely destroyed, efficient and low-delay encryption can be realized, and resource waste caused by full encryption can be avoided.

[0020] In a second aspect, the present application provides an image security transmission system for smart glasses, comprising a processor and a memory, the memory storing computer program instructions, when the computer program instructions are executed by the processor, realizing the above-mentioned image security transmission method for smart glasses.

[0021] By using the above technical solution, the above-mentioned image security transmission method for smart glasses is generated into a computer program and stored in the memory to be loaded and executed by the processor, so that a terminal device is made according to the memory and the processor, and the use is convenient.

[0022] The present application has the following beneficial effects:

[0023] The application introduces energy and channel state as a regulating lever by constructing a perceived disturbance coupling factor, amplifies the influence weight of external motion disturbance on the system when the power is reduced or the signal is deteriorated, and drives the system to make more sensitive strategy adjustment in the low resource and high risk scene; further, a dynamic balance mechanism of encryption mask threshold is constructed based on the perceived disturbance coupling factor and the current remaining power, the main control action of the power is used to automatically shrink the encryption range at low power to save energy, and the regulating action of the perceived disturbance is used to appropriately adjust the threshold when detecting violent motion to expand the encryption range, so as to prevent key information leakage caused by violent shaking of the picture; finally, combined with the energy concentration characteristics after the discrete cosine transform, the encryption mask threshold dynamically calculated is used to only encrypt the low-frequency coefficients containing the main image information, the value size and the sign bit are confused by the exclusive or operation, and the directionality is destroyed, while the image visual structure is completely destroyed to protect the core privacy security, the resource waste caused by full encryption is avoided, the adaptive adjustment of the encryption strength to the wearer's motion state and the device resource state is realized, and the device endurance time of the smart glasses is significantly prolonged. BRIEF DESCRIPTION OF DRAWINGS

[0024] The above and other objects, features and advantages of the exemplary embodiments of the present application will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0025] Figure 1 is a flow chart schematically showing an image security transmission method for smart glasses in the present application;

[0026] Figure 2 is a schematic diagram schematically showing an adaptive encryption mask threshold under the perceived disturbance and energy constraint;

[0027] Figure 3 is a performance difference comparison diagram schematically showing the present application and the prior art. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0029] The specific embodiments of the present application will be described in detail below in conjunction with the drawings.

[0030] The embodiments of the present application disclose an image security transmission method for smart glasses, referring to Figure 1This includes steps S1 to S4:

[0031] S1: Through the multi-source sensors integrated into the smart glasses, video stream data, motion posture data, and system status data are collected simultaneously.

[0032] It should be noted that due to the differences in the operating frequency and sampling clock of different sensors, the multi-source data collected cannot be strictly aligned in the time dimension, resulting in misalignment between the perceived state and the video frame. Therefore, this invention establishes a unified spatiotemporal benchmark to synchronously collect and align the features of multi-source data, thereby ensuring the accuracy of the input to the decision model.

[0033] Specifically, the smart glasses use their front-facing camera to capture real-time video streams at a preset frame rate. For each captured frame of image data, the image is converted from the RGB color space to the YCrCb color space, and the luminance component is extracted. Further processing is required because the human eye is most sensitive to changes in brightness, and encrypting the brightness components can achieve the greatest visual confusion effect at the lowest cost.

[0034] Simultaneously, the inertial measurement unit built into the smart glasses synchronously collects six-axis data from the smart glasses, including three-axis acceleration. and triaxial angular velocity These data directly reflect the wearer's head movements and overall motion status; the power management chip of the smart glasses reads the current remaining battery percentage. Read the channel received signal strength indication via the wireless communication module. .

[0035] Furthermore, spatiotemporal feature alignment is performed on the above data to ensure that the first Frame images strictly correspond to the first Motion posture data at any time, current remaining battery percentage and channel received signal strength indication .

[0036] S2: Calculate the sensing disturbance coupling factor for each frame of image based on motion posture data, current remaining battery percentage, and channel received signal strength indication.

[0037] It should be noted that simple motion data cannot fully reflect the urgency of encryption for the device under current resource constraints. This may cause the system to crash due to excessive encryption in low-power, high-risk scenarios, or provide insufficient protection against privacy leaks during vigorous movements when the power is sufficient. Therefore, this invention constructs a perceptual disturbance coupling factor, introducing energy and channel state as adjustment levers. When the power decreases or the signal deteriorates, the impact weight of external disturbances on the system is amplified, thereby driving the system to make more sensitive policy adjustments.

[0038] Specifically, according to the first Triaxial acceleration at time t Triaxial angular velocity Current remaining battery percentage and channel received signal strength indication Calculate the first The perceptual perturbation coupling factor corresponding to the frame image is calculated using the following formula:

[0039]

[0040] In the formula: For the first The perceptual perturbation coupling factor corresponding to the frame image; This is the preset scaling factor; It is the natural logarithm function; For axial index, the value is... , , ; For the first The first frame image corresponding to The absolute value of the axial angular velocity, characterizing the instantaneous rotational intensity, is normalized to... ; It is a natural constant; For the first The first frame image corresponding to The rate of change of the axial acceleration is equal to the first Frame image and the first The first frame image corresponding to The ratio of the difference in axial acceleration to the time interval between acquiring two adjacent frames of images; For the first The current remaining battery percentage at the time of frame image acquisition, with a value range of [value range missing]. ; For the first Channel received signal strength indication during frame image acquisition, normalized to .

[0041] Among them, the numerator term Characterizing the intensity of the wearer's exercise using an exponential function The rate of change of acceleration is processed because sudden shocks have a greater impact on image quality than steady acceleration. As motion intensifies, the numerator increases, leading to a higher perturbation coupling factor. Increase; denominator term Characterizing the energy and communication redundancy of the system, when the power... Reduce or signal When the denominator decreases, under the same intensity of motion, the calculated value becomes... The value increases significantly, indicating that the system's sensitivity to external motion disturbances is amplified when the power is low or the signal is weak.

[0042] Among them, the preset scaling factor Used to adjust the overall amplitude of the disturbance factor, if If the value is too small, the change in the sensing disturbance coupling factor is not significant, and it cannot drive the system out of steady state. If the scaling factor is too large, it can easily lead to system saturation and loss of sensitivity; therefore, the scaling factor should be adjusted accordingly. The range of values ​​is In this embodiment, Set as In other embodiments, implementers may set the parameters according to the actual implementation situation. .

[0043] It should be noted that this invention breaks away from the rigid mode of traditional fixed encryption strength. By establishing a perception perturbation coupling model, the encryption strength, i.e. the encryption mask threshold, can automatically decrease as the battery power decreases. In low-power scenarios, the system automatically discards the encryption of high-frequency details and only protects the core low-frequency contours, thereby significantly reducing the amount of computation and extending the battery life of smart glasses in critical moments.

[0044] S3: Calculate the encryption mask threshold for each frame of image based on the perceived perturbation coupling factor and the current remaining battery percentage.

[0045] It should be noted that since full-frame high-intensity encryption can lead to heat buildup and rapid power consumption, thereby shortening the device's battery life, this invention utilizes the energy concentration characteristics after discrete cosine transform to control the encryption frequency band range by dynamically adjusting the encryption mask threshold. This adjustment follows the power-dominated principle and the motion compensation principle, that is, when the power is low, the encryption range is automatically reduced to survive, but when violent motion is detected, the threshold is appropriately adjusted back to expand the encryption range to prevent the leakage of key information due to violent screen shaking.

[0046] Specifically, based on the first Perceptual perturbation coupling factor corresponding to the frame image and current remaining battery percentage Calculate the first The encryption mask threshold for a frame image is calculated using the following formula:

[0047]

[0048] In the formula, For the first The encryption mask threshold of the frame image represents the cutoff index of the Zigzag scan order, and its value is an integer with a range of [0,63]. This indicates the floor function; As the benchmark coefficient; For the first The current remaining battery percentage at the time of frame image acquisition is used as the main control item and the threshold is directly linearly scaled. It is a hyperbolic tangent function that maps the input to the (0,1) interval, providing smooth saturation properties; For the first The perceptual perturbation coupling factor corresponding to the frame image; This indicates taking the maximum value, used to set a minimum lower limit of 3 to ensure that even when the power is extremely low, the most core privacy, such as the DC component, is still encrypted.

[0049] Among them, the calculation formula of the encryption mask threshold constructs a dynamic balance mechanism: (1) As the main control item, it reflects the main control role of the power: when When the value decreases, the baseline value Directly reduce, force Reduced, which means that the system automatically shrinks the encryption range when the power is low, and only encrypts the most critical low-frequency coefficients, thereby greatly saving computing power and power; (2) For adjustment: when strenuous exercise occurs or the sensitivity is amplified due to low battery, Increase, making Increase, and then through multiplication. right Forward compensation, hyperbolic tangent function Utilizing its nonlinear saturation characteristics, it prevents excessive threshold expansion under extreme disturbances; this is a crisis response mechanism: even with low battery, if extreme environmental disturbances or motion risks are detected, the system will appropriately adjust the threshold and expand the encryption range to prevent scene information under violent movement, such as sudden privacy intrusions that are not effectively masked. However, when the battery is extremely low, due to... The multiplicative effect of this callback is limited, thus ensuring that the system does not run out of power due to overprotection.

[0050] Among them, modulation factor The weights used to control the impact of perceived disturbances on the threshold are as follows: If it is too small, the effect of motor compensation will not be obvious. If the modulation factor is too large, it will cause the threshold correction amplitude to be too large under low power conditions, thus negating the energy-saving effect; therefore, the modulation factor... The range of values ​​is In this embodiment, Set as In other embodiments, implementers may set the parameters according to the actual implementation situation. .

[0051] It should be noted that by introducing IMU data to construct a perception disturbance coupling factor, the system can keenly sense the wearer's violent movements. When the movement changes abruptly, the encryption mask threshold is appropriately increased through nonlinear modulation logic to prevent the leakage of key information caused by violent screen shaking or rapid scene switching, thus solving the security blind spot of the single power control strategy in dynamic scenarios.

[0052] S4: Selectively encrypt the frequency domain coefficients of each frame of the image using an encryption mask threshold and a chaotic key stream.

[0053] It should be noted that traditional encryption methods often treat all data equally, leading to wasted resources. Therefore, this invention utilizes the pseudo-randomness and sensitivity to initial values ​​of chaotic systems, combined with dynamically calculated encryption mask thresholds, to process only low-frequency coefficients containing key image information, achieving efficient and low-latency encryption.

[0054] Specifically, a chaotic key stream of the same length as the image data is generated using a pre-set chaotic system, which includes, but is not limited to, Logistic mapping and Chen's system.

[0055] Furthermore, regarding the first The luminance component of the frame image is subjected to block discrete cosine transform, with a block size of 8×8, to obtain multiple 8×8 frequency domain coefficient matrices.

[0056] Furthermore, for each frequency domain coefficient matrix, the frequency domain coefficients are traversed according to the Zigzag scan order. Under the Zigzag scan order, low-frequency coefficients are concentrated at the beginning of the sequence, containing the main contour information of the image, while high-frequency coefficients are at the end, containing detailed textures. The scan sequence number of the current frequency domain coefficient is extracted. :like The frequency domain coefficient is determined to be a key privacy coefficient. A chaotic keystream is used to perform an XOR operation and sign bit flipping on this coefficient. The XOR operation obfuscates the numerical value of the coefficient, and the sign bit flipping destroys the directionality of the coefficient, thus completely destroying the visual structure of the image. The frequency domain coefficients are determined to be non-critical high-frequency details, so they are not processed and are directly retained.

[0057] Finally, the processed frequency domain coefficients are subjected to inverse discrete cosine transform to generate encrypted video frames.

[0058] It should be noted that instead of complex full-frame pixel-level encryption, selective encryption of frequency domain coefficients is adopted. Combined with the energy concentration characteristics of Zigzag scanning, only a few low-frequency coefficients need to be XORed and flipped to achieve the effect of destroying image visibility. This greatly reduces the time complexity and computing power consumption of the algorithm, making it very suitable for embedded devices with limited computing resources, such as smart glasses.

[0059] Figure 2 This diagram illustrates the adaptive encryption mask threshold under the conditions of perceived disturbance and energy constraints. It includes three main curves: the green dashed line represents the linear decrease in remaining device power over time; the blue background waveform represents the wearer's movement intensity, with perceived disturbance sources calculated from IMU data showing significant movement peaks in the 50-80 and 140-170 frame intervals; the red solid line represents the encryption mask threshold calculated by this invention, which generally decreases with decreasing power, reflecting a low-power energy-saving strategy; however, in the intense movement interval where the blue waveform spikes, the red solid line rises instead of decreasing, showing a significant pulse-like increase. This indicates that under the influence of perceived disturbance, even with low power, once intense wearer movement is detected, the system immediately and forcibly increases the encryption bandwidth to prioritize security.

[0060] Figure 3 The graph shows a performance difference between the present invention and the prior art, comparing the energy consumption growth trends of the two technologies. In the graph, the gray dashed line represents the prior art, whose energy consumption increases rapidly in a straight line, while the orange solid line represents the present invention, whose slope gradually slows down with time and power consumption decrease. Moreover, the final cumulative value of the orange solid line is significantly lower than that of the gray dashed line, proving that the hierarchical encryption strategy of the present invention can effectively reduce computational overhead and extend device battery life.

[0061] This invention also discloses an image secure transmission system for smart glasses, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement an image secure transmission method for smart glasses according to the present invention.

[0062] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. An image security transmission method for smart glasses, characterized in that, The method comprises the following steps: Synchronously collecting video stream data, motion posture data and system state data through multi-source sensors integrated in smart glasses, and aligning the collected data in time and space characteristics; According to the motion posture data, the current percentage of remaining power in the system state data and the channel received signal strength indication, calculating the perception disturbance coupling factor corresponding to each frame of image; Based on the perception disturbance coupling factor and the current percentage of remaining power, calculating the encryption mask threshold of each frame of image; Using the encryption mask threshold and the preset chaotic key stream, selectively encrypting the frequency domain coefficients of each frame of image to obtain the encrypted video frame.

2. The image security transmission method for smart glasses according to claim 1, characterized in that, The synchronously collecting video stream data, motion posture data and system state data through multi-source sensors integrated in smart glasses comprises the following steps: Collecting real-time video stream through the front camera of smart glasses at a preset frame rate, and converting each frame of image data collected from RGB color space to YCrCb color space to extract the luminance component Y; Collecting three-axis acceleration and three-axis angular velocity through the inertial measurement unit built-in smart glasses; Reading the current percentage of remaining power through the power management chip of smart glasses, and reading the channel received signal strength indication through the wireless communication module.

3. The image security transmission method for smart glasses according to claim 1, characterized in that, The calculation formula of the perception disturbance coupling factor corresponding to each frame of image is: ； In the formula: is the first frame image corresponding to the first axis angular velocity; is a natural logarithm function; is an axial index, taking values , , ; is the first frame image corresponding to the first axis angular velocity absolute value; is a natural constant; is the first frame image corresponding to the first axis acceleration rate of change; is the first frame image acquisition time of the current remaining percentage of power; is the first frame image acquisition time of the channel received signal strength indication.

4. The image security transmission method for smart glasses according to claim 3, characterized in that, The first Rate of change of the axis acceleration Equal to the first The first The axis acceleration corresponding to the first The first The difference value of the axis acceleration corresponding to the first 5. The image security transmission method for smart glasses according to claim 1, characterized in that, The calculation formula of the encryption mask threshold is: ； In the formula, is the first encrypted mask threshold of the frame image; represents a floor operation; is a reference coefficient; is the first current remaining percentage of power when the frame image is collected; is a hyperbolic tangent function; is the first perception disturbance coupling factor corresponding to the frame image; represents taking the maximum value.

6. The image security transmission method for smart glasses according to claim 1, wherein, The selectively encrypting the frequency domain coefficients of each frame of image using the encryption mask threshold and the preset chaotic key stream comprises the following steps: Generating a chaotic key stream with the same length as the image data using a preset chaotic system; The method comprises the following steps: The luminance component of the frame image is block discrete cosine transformed to obtain a plurality of frequency domain coefficient matrices; For each frequency domain coefficient matrix, traversing the frequency domain coefficients in Zigzag scanning order, and selectively encrypting the frequency domain coefficients of each frame of image.

7. The image security transmission method for smart glasses according to claim 6, characterized in that, The selectively encrypting the frequency domain coefficients of each frame of image comprises the following steps: For the frequency domain coefficients determined as key privacy coefficients, performing XOR operation and symbol bit flipping operation using the chaotic key stream; For the frequency domain coefficients determined as non-key high-frequency details, keeping the numerical values unchanged; Performing inverse discrete cosine transform on the processed frequency domain coefficients to generate the encrypted video frame.

8. The image security transmission method for smart glasses according to claim 7, characterized in that, The determination method of the frequency domain coefficients comprises the following steps: Extracting the scan order number of the current frequency domain coefficient; If the scan order number is less than the encryption mask threshold, determining that the frequency domain coefficient is a key privacy coefficient; If the scan order number is greater than or equal to the encryption mask threshold, determining that the frequency domain coefficient is a non-key high-frequency detail.

9. An image security transmission system for smart glasses, characterized in that, The method comprises the following steps: A processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, realizing the image security transmission method for smart glasses according to any one of claims 1-8.

Images