Image acquisition system and method, wearable device

Abstract

The present disclosure belongs to the technical field of intelligent wearing, and provides an image acquisition system and method, and a wearable device. The image acquisition system comprises an image acquisition module, a first processing module, a second processing module and a third processing module. The image acquisition module acquires image data at a first frame rate, and acquires image data at a second frame rate in response to acquisition of a high-frame acquisition instruction. The second frame rate is greater than the first frame rate. The first processing module processes adjacent two frames of image data using a first preset algorithm, and wakes up the second processing module when a first processing result meeting a first preset condition is acquired. The third processing module is woken up when a second processing result meets a second preset condition. The second processing module processes two frames of target image data using a second preset algorithm, and feeds back the second processing result to the first processing module. The third processing module sends a high-frame acquisition instruction to the image acquisition module. The video frame rate can be switched according to the video recording scene, and the performance and power consumption are balanced.

Description

Technical Field

[0001] This disclosure belongs to the field of smart wearable technology, and in particular relates to an image acquisition system and method, and a wearable device. Background Technology

[0002] With the increasing popularity of wearable devices, these devices have been applied to various daily and sports scenarios. Taking smart glasses as an example, wearers can collect image data during daily travel and leisure activities. Currently, consumer-grade smart glasses are generally equipped with multi-core system-on-a-chip computing platforms, which can support users to continuously carry out recording tasks in various scenarios such as life recording and outdoor activities.

[0003] However, most existing smart glasses recording solutions can only achieve basic image capture and calculation output at a fixed frame rate, and cannot dynamically adjust computing power to balance performance and power consumption. Summary of the Invention

[0004] To address the aforementioned issues, this disclosure provides an image acquisition system and method, as well as a wearable device.

[0005] This disclosure is achieved through the following technical solution: In a first aspect, embodiments of this disclosure provide an image acquisition system. The system includes: an image acquisition module, a first processing module, a second processing module, and a third processing module; The image acquisition module is configured to: acquire image data at a first frame rate, and acquire image data at a second frame rate in response to the acquisition of a high frame acquisition command; wherein the second frame rate is greater than the first frame rate; The first processing module is configured to: read image data at a first preset period, process two adjacent frames of image data using a first preset algorithm, and wake up the second processing module when a first processing result that meets the first preset condition is obtained; obtain a second processing result, and wake up the third processing module when the second processing result meets the second preset condition; The second processing module is configured to: after being woken up by the first processing module, read two frames of target image data, process the two frames of target image data using a second preset algorithm, obtain the second processing result, and feed the second processing result back to the first processing module; The third processing module is configured to send a high-frame acquisition command to the image acquisition module after being woken up by the first processing module.

[0006] Furthermore, The first preset period is the same as the frame interval of the first frame rate, and the first preset algorithm includes the frame difference method; the system further includes: a timing module; The timing module is configured to continuously send trigger commands to the first processing module at the first preset period; The first processing module is further configured to: in response to the acquisition of the trigger instruction, read image data sequentially; after reading two adjacent frames of image data, compare the image features of the two adjacent frames of image data and obtain a comparison result; after obtaining the comparison result that meets the third preset condition, determine the next frame of image data in the two adjacent frames of image data corresponding to the comparison result that meets the third preset condition as the reference frame image data; process the next frame of image data read sequentially after the reference frame image data and the reference frame image data using the first preset algorithm to obtain the first processing result, and determine the next frame of image data as the new reference frame image data; After obtaining the first processing result that meets the first preset condition, a first wake-up command is sent to the second processing module.

[0007] Furthermore, The second preset algorithm includes a lightweight convolutional neural network; the second processing module is further configured to: in response to the acquisition of the first wake-up command, read two frames of target image data, process the two frames of target image data using the lightweight convolutional neural network; acquire a second processing result, feed the second processing result back to the first processing module, and then enter a sleep state.

[0008] Furthermore, The first processing module is further configured to: directly enter a sleep state when only the first frame of image data is read; and enter a sleep state after obtaining the comparison result or the first processing result.

[0009] Furthermore, The first processing module is further configured to: after obtaining a second processing result that meets the second preset condition, send a second wake-up command to the third processing module; and in response to obtaining a frame rate switching command, send a low frame acquisition command to the image acquisition module. The second processing module is further configured to: in response to the acquisition of a third wake-up command, read display image data, process the display image data using a third preset algorithm, obtain a third processing result, and feed the third processing result back to the third processing module; after the third processing result is fed back, enter a sleep state; wherein, the third preset algorithm includes a single-stage target detection algorithm; The third processing module is further configured to: in response to the second wake-up command, send the high frame rate acquisition command to the image acquisition module; read image data at a second preset period, and after reading the display image data, send the third wake-up command to the second processing module; wherein the second preset period is the same as the frame interval of the second frame rate; based on the third processing result, determine the target information to be displayed, and if there is no target information, send the frame rate switching command to the first processing module and enter a sleep state; The image acquisition module is further configured to: acquire image data at the first frame rate in response to the acquisition of the low frame acquisition command.

[0010] Furthermore, The system also includes: a fourth processing module; The fourth processing module is configured to: in response to the acquisition of the sharing start command, initiate sharing transmission at a third preset period and send the shared image data to the target device through the wireless network; after the shared image data is sent, enter a sleep state and wait to start sharing transmission again.

[0011] Furthermore, The first processing module is further configured to: in response to the acquisition of an image preprocessing instruction, read the image data to be shared, preprocess the image data to be shared, acquire the shared image data, write the shared image data into a target storage area; send a preprocessing completion instruction to the fourth processing module, and enter sleep mode; wherein, the preprocessing includes compression and encryption; The fourth processing module is further configured to: after initiating shared transmission, send the image preprocessing instruction to the first processing module; in response to the acquisition of the preprocessing completion instruction, read the shared image data from the target storage area, and send the shared image data to the target device through the wireless network within a preset sending window; wherein the duration of the third preset period is greater than the duration of the preset sending window.

[0012] Secondly, based on the same inventive concept, embodiments of this disclosure also provide an image acquisition method, applied to an image acquisition system including an image acquisition module, a first processing module, a second processing module, and a third processing module; the method includes: The image acquisition module acquires image data at a first frame rate; The first processing module reads image data at a first preset period, processes two adjacent frames of image data using a first preset algorithm, and wakes up the second processing module when it obtains a first processing result that meets the first preset condition. After being awakened by the first processing module, the second processing module reads two frames of target image data, processes the two frames of target image data using a second preset algorithm, obtains a second processing result, and feeds the second processing result back to the first processing module. The first processing module obtains the second processing result, and wakes up the third processing module when the second processing result meets the second preset condition; After being awakened by the first processing module, the third processing module sends a high-frame acquisition command to the image acquisition module. The image acquisition module responds to the acquisition of the high frame acquisition command and acquires image data at a second frame rate; wherein the second frame rate is greater than the first frame rate.

[0013] Furthermore, The image acquisition system further includes: a fourth processing module; the method further includes: In response to the acquisition of the sharing start command, the fourth processing module starts the sharing transmission at a third preset cycle and sends the shared image data to the target device through the wireless network; After the shared image data is sent, it enters a sleep state, waiting to restart the shared transmission.

[0014] Thirdly, based on the same inventive concept, embodiments of this disclosure also provide a wearable device, which includes any of the aforementioned image acquisition systems.

[0015] The image acquisition system in this embodiment of the present disclosure achieves the switching of the recording frame rate according to the recording scene during the recording process through the step-by-step wake-up and collaborative processing of the first processing module, the second processing module and the third processing module, thereby adjusting the computing power and balancing the performance and power consumption during the image acquisition process.

[0016] Other features and advantages of this disclosure will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the disclosure. The objects and other advantages of this disclosure may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A block diagram of an image acquisition system provided in this disclosure embodiment; Figure 2 This is a flowchart of an image acquisition method provided in an embodiment of the present disclosure. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0020] Firstly, Figure 1 This is a block diagram of an image acquisition system provided in an embodiment of the present disclosure, such as... Figure 1 As shown in the figure, this disclosure provides an image acquisition system, which includes an image acquisition module, a first processing module, a second processing module, and a third processing module.

[0021] The image acquisition module is configured to acquire image data at a first frame rate and, in response to a high frame rate acquisition command, acquire image data at a second frame rate, wherein the second frame rate is greater than the first frame rate. The image acquisition module is used to acquire images, convert the acquired images into image data, and store the image data in a designated storage area. The image acquisition module includes at least two image acquisition frame rates: the first frame rate is a low frame rate with low power consumption, and the second frame rate is a high frame rate with higher power consumption.

[0022] The first processing module is configured to: read image data at a first preset period, process two adjacent frames of image data using a first preset algorithm, and wake up the second processing module when a first processing result that meets the first preset condition is obtained; acquire a second processing result, and wake up the third processing module when the second processing result meets the second preset condition. The first processing module can be a low-power processing core, which enables the image acquisition module to acquire images at a first frame rate and can read image data from a specified storage area and process the image data; at the same time, it can also wake up the second processing module based on the first processing result and wake up the third processing module based on the second processing result.

[0023] The second processing module is configured to: upon being awakened by the first processing module, read two frames of target image data, process the two frames of target image data using a second preset algorithm, obtain a second processing result, and feed the second processing result back to the first processing module. After being awakened by the first processing module, the second processing module begins operation, retrieving the latest two frames of image data from a designated storage area in chronological order as target image data. After retrieving two adjacent frames of target image data in chronological order, it processes the target image data; the target image data can be the two frames of image data corresponding to the first processing result that meet the first preset condition. The target image data can also be the latest frame of image data when the second processing module is awakened by the first processing module and the next frame of image data written to the designated storage area.

[0024] It should be noted that the target image data can also be two frames of image data corresponding to the first processing result that meet the first preset condition. After the first processing module wakes up the second processing module, the data is sent directly from the first processing module to the second processing module. Whether the processing modules directly transmit image data depends on the amount of image data. If the amount of image data is large, the processing module reads it from a designated storage area. During reading, the corresponding image data can be located in the designated storage module using preset identifiers, such as frame number or storage address.

[0025] The third processing module is configured to send a high-frame-rate acquisition command to the image acquisition module after being woken up by the first processing module. Typically, the third processing module is a high-performance processing core, and its power consumption may be greater than that of the first processing module. The third processing module will only be woken up after the second processing module has processed the target image data and the data meets the second preset condition, causing the image acquisition module to switch to a high frame rate for image acquisition.

[0026] The image acquisition system in this embodiment includes at least an image acquisition module, a first processing module, a second processing module, and a third processing module. The image acquisition module acquires images at a first frame rate and converts them into image data. The first processing module processes two adjacent frames of image data using a first preset algorithm. When the first processing result meets a first preset condition, the first processing module wakes up the second processing module for further processing. After being woken up by the first processing module, the second processing module reads two frames of target image data, processes the target image data using a second preset algorithm, and feeds back the second processing result to the first processing module. The first processing module obtains the second processing result, and when the second processing result meets a second preset condition, wakes up the third processing module. After being woken up by the first processing module, the third processing module sends a high-frame acquisition command to the image acquisition module. At this time, the image acquisition module switches from the first frame rate to the second frame rate. Through the step-by-step wake-up and collaborative processing of the first, second, and third processing modules, the wearable device can switch the recording frame rate according to the recording scene during recording, thereby adjusting computing power and balancing performance and power consumption during image acquisition.

[0027] The image acquisition system in this embodiment, when applied to a wearable device, can enable the wearable device to switch the recording frame rate according to the recording scene during recording, balancing the performance and power consumption of the wearable device's recording function, thereby improving the wearable device's battery life and meeting the user's actual needs for all-day use. The following description uses an image acquisition system configured in a wearable device as an example.

[0028] In some examples, a Cortex-M55-HE processor can be used as the first processing module, an Ethos-U85 micro neural network processing unit or an Ethos-U55 micro neural network processing unit as the second processing module, and a Cortex-A55 processor as the third processing module. The first, second, and third processing modules can be adjusted according to the actual situation and needs of the wearable device. Typically, the first processing module uses a lower-power module to effectively control the overall power consumption of the wearable device.

[0029] Furthermore, the image acquisition module includes a camera, an image signal processor, and a MIPI interface (Mobile Industry Processor Interface). The camera is used to acquire optical images and convert them into raw image data. The raw image data is output to the image signal processor through the MIPI interface. The image signal processor is used to perform image processing on the raw image data. Image processing includes, but is not limited to, at least one of noise reduction, white balance, exposure correction, and color enhancement. The processed image data is output through the MIPI interface and stored in a designated storage area. Alternatively, it can be configured to output directly to the first processing module, but considering that the first processing module will enter a sleep state, it is preferred to store it in the designated storage area. The image acquisition module supports two acquisition modes: a first frame rate and a second frame rate. The first frame rate is a low frame rate mode, and the second frame rate is a high frame rate mode. For example, the first frame rate can be 30 Hz, that is, an image is acquired every 33 milliseconds, and the second frame rate can be 60 Hz, that is, an image is acquired every 16 milliseconds. When the image acquisition module receives a high-frame acquisition command from the third processing module, it switches from the first frame rate to the second frame rate for image data acquisition. Simultaneously, the data path of the MIPI interface switches from the first processing module side to the third processing module side. The image data acquired at the second frame rate is output through the MIPI interface and stored in a designated storage area. The third processing module then reads and processes this data from the designated storage area. The high-frame acquisition command is a control signal used to trigger the frame rate switching of the image acquisition module and the data path switching of the MIPI interface. The MIPI interface serves as the image data transmission interface, used to transmit the image data output by the image acquisition module. The camera and image signal processor can be integrated into the same chip or are independent devices, depending on the actual situation and needs of the wearable device. The designated storage area can be any available volatile or non-volatile storage area in the wearable device or image acquisition system, used to cache or store the image data output by the image acquisition module. The specific selection and setting of the designated storage area are not specifically limited here.

[0030] In some examples, the image acquisition system also includes a timing module; the timing module is configured to continuously send trigger commands to the first processing module at a first preset period. The timing module can be integrated into a normally open power domain, which is used to continuously power the timing module and other modules that need to operate continuously. The normally open power domain is usually kept continuously powered.

[0031] The first processing module is further configured to: in response to the acquisition of a trigger command, read image data sequentially; after reading two adjacent frames of image data, compare the image features of the two adjacent frames of image data and obtain a comparison result; after obtaining a comparison result that meets a third preset condition, determine the next frame of image data in the two adjacent frames of image data corresponding to the comparison result that meets the third preset condition as the reference frame image data; process the next frame of image data read sequentially after the reference frame image data and the reference frame image data using a first preset algorithm to obtain a first processing result, and determine the next frame of image data as the new reference frame image data; after obtaining the first processing result that meets the first preset condition, send a first wake-up command to the second processing module.

[0032] In this embodiment, the wearable device can be activated by the user to start the recording mode, thereby waking up the image acquisition system; or the control system in the wearable device can be activated by the user to start the relevant function. After the image acquisition system is woken up, the image acquisition module acquires images at a first frame rate, converts them into image data, and transmits the image data to a designated storage area. The timing module continuously sends trigger commands to the first processing module at a first preset period. In response to the trigger commands, the first processing module reads the first frame of image data from the designated storage area. At this time, only the first frame of image data is read, and no other processing is required. After reading the first frame of image data, the system directly enters a sleep state, reducing the power consumption of the image acquisition system. The first preset period can be the same as the frame interval of the first frame rate. For example, when the first frame rate is 30 Hz, the first preset period corresponds to 33 milliseconds. After the first preset period has elapsed, the timing module sends a trigger command to the first processing module, and the first processing module is woken up again, reading two frames of image data from the designated storage area. At this time, the image features of two adjacent frames of image data are compared, that is, the image features of the first frame of image data and the second frame of image data are compared to obtain a comparison result. Image feature comparison includes at least one of block mean comparison, grayscale histogram comparison, and downsampled pixel difference comparison. Image features are compared between two adjacent frames of image data to obtain image feature differences. When the image feature difference between two adjacent frames exceeds a first preset threshold, it is determined that motion may exist. The image feature difference is used as the comparison result; if the image feature difference exceeds the first preset threshold, it meets a third preset condition. If the comparison result of the first and second frames of image data meets the third preset condition, subsequent steps are performed. If the comparison result does not meet the third preset condition, the first processing module enters a sleep state, waiting for the next trigger instruction to compare the second and third frames of image data again and obtain the comparison result. This process continues until a comparison result that meets the third preset condition is obtained, at which point subsequent steps begin.

[0033] The next frame of image data in the two adjacent frames corresponding to the comparison result that meets the third preset condition is determined as the reference frame image data. Taking the first frame image data and the second frame image data as examples, the second frame image data is the reference frame image data. After obtaining a comparison result, the first processing module enters a sleep state, waiting for the next trigger command to wake it up again. After being woken up by the trigger command again, it reads the reference frame image data (the second frame image data) and the next frame image data (the third frame image data) of the reference frame image data. The first preset algorithm is used to process the second frame image data and the third frame image data to obtain the first processing result. At this time, the next frame image data (the third frame image data) is determined as the new reference frame image data (the third frame image data). After obtaining the first processing result, the first processing module enters a sleep state. The first preset algorithm includes the frame difference method. The first processing result is the processing result of processing two adjacent frames of image data using the frame difference method. The frame difference method determines whether there is motion in the scene by comparing the differences of corresponding pixels in two consecutive frames of image data. The grayscale value of each pixel in the previous frame image data and the grayscale value of the corresponding pixel in the next frame image data are obtained. The absolute value of the difference between the grayscale values ​​of the two frames at the same pixel is calculated. When the absolute value of the difference is greater than a second preset threshold, the pixel is marked as a moving pixel; otherwise, it is marked as a stationary pixel. The second preset threshold is a pre-set grayscale difference judgment threshold. The marking results of all pixels are statistically analyzed to form a binary motion mask, in which moving pixels are marked as 1 and stationary pixels are marked as 0. Further, the total number of moving pixels marked as 1 in the binary motion mask is counted as the motion value. When the motion value is greater than a third preset threshold, it is determined that the first preset condition is met, and it is determined that there is a moving target in the two adjacent frame image data. After obtaining the first processing result that meets the first preset condition, a first wake-up command is sent to the second processing module.

[0034] In some instances, the second preset algorithm includes a lightweight convolutional neural network; the second processing module is further configured to: in response to a first wake-up command, process two frames of target image data using the lightweight convolutional neural network; obtain a second processing result, feed the second processing result back to the first processing module, and then enter a sleep state. A lightweight convolutional neural network refers to a network with few parameters, low computational cost, and suitability for operation on devices with limited computing and energy resources. For example, a convolutional neural network model running on a wearable device can be a lightweight convolutional neural network, built upon depthwise separable convolution. Depthwise separable convolution decomposes standard convolution into two steps: depthwise convolution and pointwise convolution. Depthwise convolution performs convolution on each input channel using a separate kernel, while pointwise convolution uses a 1×1 convolution to fuse the channels of the depthwise convolution output. Compared to standard convolution, depthwise separable convolution significantly reduces computational cost. Before using a lightweight convolutional neural network to process two frames of target image data, the target image data can be preprocessed, for example, by downsampling the image to 96×96 resolution and converting it to a grayscale image.

[0035] This lightweight convolutional neural network is a binary classification model for motion and stillness, used to determine whether the motion corresponding to two adjacent frames of target image data is real motion or pseudo-motion. For example, it can determine whether the motion is real or pseudo-motion caused by changes in light, swaying leaves, or undulating water plants. The input to this binary classification model is a 96×96 grayscale image. The network structure includes 2 to 3 depthwise separable convolutional layers and fully connected layers, with fewer than 100,000 parameters, fewer than 5 million multiply-accumulate operations, a model size of less than 200 kilobytes, an inference time of 2 to 4 milliseconds, and an accuracy greater than 95%. The second processing module uses this lightweight convolutional neural network to process the two frames of target image data and obtain a second processing result. This second processing result indicates whether a real moving target exists in the two frames of target image data; when the second processing result indicates the presence of a real moving target, it meets the second preset condition.

[0036] In some examples, the first processing module is also configured to: after obtaining the second processing result that meets the second preset condition, send a second wake-up command to the third processing module, that is, when the second preset algorithm determines that the motion corresponding to the two frames of target image data is real motion, send a second wake-up command to the third processing module.

[0037] The third processing module is also configured to: in response to the second wake-up command, send a high-frame acquisition command to the image acquisition module; read image data at a second preset period; and after reading the display image data, send a third wake-up command to the second processing module; wherein the second preset period is the same as the frame interval of the second frame rate, for example, if the second preset period is 16 milliseconds and the second frame rate is 60 Hz, the frame interval is 16 milliseconds. Display image data can be read from a specified storage area. The display image data is one frame of image data that needs to be displayed to the user, and its determination method can be through preset identifiers, such as frame number, storage address, etc. It should be noted that after the third processing module reads the display image data, it can write the display image data into its own buffer or register.

[0038] After the third processing module is awakened by the second wake-up command, it can send a signal to the timing module to switch the timing module from the first preset period to the second preset period. After the third processing module finishes reading a frame of image data and sends the third wake-up command to the second processing module, it can enter a sleep state and wait for the timing module to wake it up by the trigger command at the second preset period and read the image data again. Since the third processing module corresponds to the high frame rate of the second frame rate, the third processing module can also be in a long-term power-on state.

[0039] The second processing module is also configured to: in response to the acquisition of the third wake-up command, read the display image data, process the display image data using a third preset algorithm, obtain the third processing result, and feed the third processing result back to the third processing module; after the third processing result is fed back, it enters a sleep state; wherein, the third preset algorithm includes a single-stage target detection algorithm, such as the TinyYOLO algorithm or the YOLO algorithm. The third processing result is used to determine whether the display image data includes target information that needs to be displayed to the user. Typically, considering that the amount of display image data may be large, the second processing module reads the display image data from a designated storage area, which can be determined by preset identifiers, such as frame number or storage address. It should be noted that the third processing module can also directly send the display image data to the second processing module, depending on the number of image frames and the amount of data in the display image data.

[0040] The third processing module is also configured to: determine the target information to be displayed based on the third processing result; if no target information exists, send a frame rate switching command to the first processing module and enter a sleep state. After the third processing module sends the frame switching command to the first processing module, it can enter a sleep state to reduce power consumption. Based on the third processing result, if the target information exists, it can be displayed on the wearable device.

[0041] It should be noted that when the third processing module needs to display the target information on the wearable device, it can read the display image data again from the designated storage area, or it can directly read the previously acquired display image data from its own register or buffer; the display of the target information on the wearable device also requires the participation of the display control function module.

[0042] The first processing module is further configured to: in response to the acquisition of a frame rate switching command, send a low-frame acquisition command to the image acquisition module. When the image acquisition module receives the low-frame acquisition command sent by the third processing module, the image acquisition module switches from the second frame rate to the first frame rate for image data acquisition. Simultaneously, the data path of the MIPI interface switches from the third processing module side to the first processing module side. The image data acquired at the first frame rate is output through the MIPI interface and stored in a designated storage area, which is then read and processed by the first processing module. The low-frame acquisition command is a control signal used to trigger the frame rate switching of the image acquisition module and the data path switching of the MIPI interface. The MIPI interface serves as an image data transmission interface for transmitting the image data output by the image acquisition module. After being awakened by the frame rate switching command, the first processing module can send a signal to the timing module, causing the timing module to switch from a second preset period to a first preset period.

[0043] In some examples, the image acquisition system further includes a fourth processing module; the fourth processing module is configured to: in response to the acquisition of a sharing start command, initiate sharing transmission at a third preset period, and send shared image data to the target device via a wireless network; after the shared image data transmission is completed, enter a sleep state, waiting to restart sharing transmission. After the user activates the sharing mode, the wearable device's control system sends a sharing start command to the fourth processing module. The third preset period is the cycle for performing one sharing transmission. After completing one sharing transmission, the system enters a sleep state, waiting for the third preset period to restart sharing transmission.

[0044] Furthermore, the fourth processing module is also configured to send image preprocessing instructions to the first processing module after the shared transmission is initiated.

[0045] The first processing module is also configured to: in response to the acquisition of an image preprocessing instruction, read the image data to be shared, preprocess the image data to be shared, acquire the shared image data, and write the shared image data to the target storage area; wherein, the preprocessing includes compression and encryption. The image data to be shared can be read from a specified storage area. The image data to be shared is a frame of image data waiting to be sent to the target device, and its determination method can be through preset identifiers, such as frame number, storage address, etc. The target storage area can be a partition in a specified storage area storing image data, or a buffer set up by the image acquisition system; no further specific limitations are made here. After completing the writing of the shared image data to the target storage area, a preprocessing completion instruction is sent back to the fourth processing module, and the module enters sleep mode. It should be noted that the shared image data can also be directly sent to the fourth processing module, depending on the number of image frames and the amount of data in the shared image data.

[0046] The fourth processing module is also configured to: in response to the acquisition of a preprocessing completion command, read shared image data from the target storage area and transmit the shared image data to the target device via a wireless network within a preset transmission window; wherein the duration of the third preset period is longer than the duration of the preset transmission window. The preset transmission window is a time window in which shared image data can be output to the target device via a wireless network. When reading shared image data from the target storage area, the shared image data can be identified by preset identifiers, such as frame sequence number, storage address, etc.

[0047] In this embodiment, in the sharing mode, the first processing module compresses and encrypts the image data to be shared stored in the designated storage area to generate shared image data, writes the shared image data to the target storage module, and the fourth processing module reads the shared image data from the target storage module and sends it to the target device via a wireless network protocol. Compression reduces the amount of image data, lowering transmission bandwidth and power consumption. Encryption protects the image data by encoding, preventing unauthorized devices from acquiring or tampering with it during transmission. Compression and encryption can be performed sequentially, or other processing orders can be used according to actual needs. Shared transmission is completed within a preset sending window, which can be set to 8 milliseconds, and the third preset period can be set to 100 milliseconds. The image data to be shared can be one frame or multiple frames; correspondingly, the shared image data obtained after preprocessing can also be one frame or multiple frames. Taking a 480×480 resolution image data as an example, after compression, the transmission time of 6 frames of shared image data under low-speed WiFi4 network (IEEE 802.11n standard wireless network) is approximately 8 milliseconds. Therefore, setting the preset sending window to 8 milliseconds is sufficient to meet the transmission requirements. After the current frame of shared image data is transmitted, the fourth processing module enters a sleep state. It is woken up again after a third preset period to execute the shared transmission of the next frame of shared image data. The fourth processing module can be woken up by its own timer or by an external timer; no specific limitation is made here. In this embodiment, a Cortex-M33 processor can be used as the fourth processing module.

[0048] Secondly, Figure 2 This is a flowchart of an image acquisition method provided in an embodiment of the present disclosure, such as... Figure 2 As shown, based on the same inventive concept, this disclosure also provides an image acquisition method, applied to an image acquisition system including an image acquisition module, a first processing module, a second processing module, and a third processing module; the image acquisition method includes: S1: The image acquisition module acquires image data at the first frame rate; S2: The first processing module reads image data at a first preset period, processes adjacent two frames of image data using a first preset algorithm, and wakes up the second processing module when it obtains a first processing result that meets the first preset condition. S3: After being awakened by the first processing module, the second processing module reads two frames of target image data, processes the two frames of target image data using the second preset algorithm, obtains the second processing result, and feeds the second processing result back to the first processing module. S4: The first processing module obtains the second processing result, and wakes up the third processing module when the second processing result meets the second preset condition; S5: After being awakened by the first processing module, the third processing module sends a high-frame acquisition command to the image acquisition module; S6: The image acquisition module responds to the high frame acquisition command and acquires image data at a second frame rate; wherein the second frame rate is greater than the first frame rate.

[0049] In some examples, the recording system also includes: a fourth processing module; the recording method also includes: S7: In response to the acquisition of the sharing start command, the fourth processing module starts the sharing transmission at the third preset cycle and sends the shared image data to the target device through the wireless network; S8; After the shared image data is sent, it enters a sleep state, waiting to restart the shared transmission.

[0050] Thirdly, based on the same inventive concept, embodiments of this disclosure also provide a wearable device, which includes any of the aforementioned image acquisition systems. The wearable device can be a smartwatch, smart bracelet, smart glasses, smart helmet, or other device with image acquisition capabilities.

[0051] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. An image acquisition system, characterized in that, The system includes: an image acquisition module, a first processing module, a second processing module, and a third processing module; The image acquisition module is configured to: acquire image data at a first frame rate, and acquire image data at a second frame rate in response to the acquisition of a high frame acquisition command; wherein the second frame rate is greater than the first frame rate; The first processing module is configured to: read image data at a first preset period, process two adjacent frames of image data using a first preset algorithm, and wake up the second processing module when a first processing result that meets the first preset condition is obtained; obtain a second processing result, and wake up the third processing module when the second processing result meets the second preset condition; The second processing module is configured to: after being woken up by the first processing module, read two frames of target image data, process the two frames of target image data using a second preset algorithm, obtain the second processing result, and feed the second processing result back to the first processing module; The third processing module is configured to send a high-frame acquisition command to the image acquisition module after being woken up by the first processing module.

2. The system according to claim 1, characterized in that, The first preset period is the same as the frame interval of the first frame rate, and the first preset algorithm includes the frame difference method; the system further includes: a timing module; The timing module is configured to continuously send trigger commands to the first processing module at the first preset period; The first processing module is further configured to: in response to the acquisition of the trigger instruction, read image data sequentially; after reading two adjacent frames of image data, compare the image features of the two adjacent frames of image data and obtain a comparison result; after obtaining the comparison result that meets the third preset condition, determine the next frame of image data in the two adjacent frames of image data corresponding to the comparison result that meets the third preset condition as the reference frame image data; process the next frame of image data read sequentially after the reference frame image data and the reference frame image data using the first preset algorithm to obtain the first processing result, and determine the next frame of image data as the new reference frame image data; After obtaining the first processing result that meets the first preset condition, a first wake-up command is sent to the second processing module.

3. The system according to claim 2, characterized in that, The second preset algorithm includes a lightweight convolutional neural network; the second processing module is further configured to: in response to the acquisition of the first wake-up command, read two frames of target image data, process the two frames of target image data using the lightweight convolutional neural network; acquire a second processing result, feed the second processing result back to the first processing module, and then enter a sleep state.

4. The system according to claim 2, characterized in that, The first processing module is further configured to: directly enter a sleep state when only the first frame of image data is read; and enter a sleep state after obtaining the comparison result or the first processing result.

5. The system according to claim 1, characterized in that, The first processing module is further configured to: after obtaining a second processing result that meets the second preset condition, send a second wake-up command to the third processing module; and in response to obtaining a frame rate switching command, send a low frame acquisition command to the image acquisition module. The second processing module is further configured to: in response to the acquisition of a third wake-up command, read display image data, process the display image data using a third preset algorithm, obtain a third processing result, and feed the third processing result back to the third processing module; after the third processing result is fed back, enter a sleep state; wherein, the third preset algorithm includes a single-stage target detection algorithm; The third processing module is further configured to: in response to the second wake-up command, send the high frame rate acquisition command to the image acquisition module; read image data at a second preset period, and after reading the display image data, send the third wake-up command to the second processing module; wherein the second preset period is the same as the frame interval of the second frame rate; based on the third processing result, determine the target information to be displayed, and if there is no target information, send the frame rate switching command to the first processing module and enter a sleep state; The image acquisition module is further configured to: acquire image data at the first frame rate in response to the acquisition of the low frame acquisition command.

6. The system according to claim 1, characterized in that, The system also includes: a fourth processing module; The fourth processing module is configured to: in response to the acquisition of the sharing start command, initiate sharing transmission at a third preset period and send the shared image data to the target device through the wireless network; after the shared image data is sent, enter a sleep state and wait to start sharing transmission again.

7. The system according to claim 6, characterized in that, The first processing module is further configured to: in response to the acquisition of an image preprocessing instruction, read the image data to be shared, preprocess the image data to be shared, acquire the shared image data, and write the shared image data into a target storage area; The preprocessing is completed and a preprocessing instruction is sent to the fourth processing module, and then the system enters sleep mode; wherein, the preprocessing includes compression and encryption; The fourth processing module is further configured to: after initiating shared transmission, send the image preprocessing instruction to the first processing module; in response to the acquisition of the preprocessing completion instruction, read the shared image data from the target storage area, and send the shared image data to the target device through the wireless network within a preset sending window; wherein the duration of the third preset period is greater than the duration of the preset sending window.

8. An image acquisition method, characterized in that, The method is applied to an image acquisition system comprising an image acquisition module, a first processing module, a second processing module, and a third processing module; the method includes: The image acquisition module acquires image data at a first frame rate; The first processing module reads image data at a first preset period, processes two adjacent frames of image data using a first preset algorithm, and wakes up the second processing module when it obtains a first processing result that meets the first preset condition. After being awakened by the first processing module, the second processing module reads two frames of target image data, processes the two frames of target image data using a second preset algorithm, obtains a second processing result, and feeds the second processing result back to the first processing module. The first processing module obtains the second processing result, and wakes up the third processing module when the second processing result meets the second preset condition; After being awakened by the first processing module, the third processing module sends a high-frame acquisition command to the image acquisition module. The image acquisition module responds to the acquisition of the high frame acquisition command and acquires image data at a second frame rate; wherein the second frame rate is greater than the first frame rate.

9. The method according to claim 8, characterized in that, The image acquisition system further includes: a fourth processing module; the method further includes: In response to the acquisition of the sharing start command, the fourth processing module starts the sharing transmission at a third preset cycle and sends the shared image data to the target device through the wireless network; After the shared image data is sent, it enters a sleep state, waiting to restart the shared transmission.

10. A wearable device, characterized in that, Includes the image acquisition system described in any one of claims 1-7.

Images