Image processing method, model training method, device and medium

Abstract

This disclosure provides an image processing method, model training method, device, and medium. The method includes: acquiring an image to be processed; inputting the image to be processed into a first target image processing model to obtain a first target feature image output by the first target image processing model; subtracting the first target feature image from the image to be processed to obtain a target preprocessed image; inputting the target preprocessed image into a second target image processing model to obtain a second target feature image output by the second target image processing model; and adding the first target feature image and the second target feature image to obtain a target image. The target image extracted by this technical solution can include both primary and secondary features from the image to be processed, thereby ensuring that the image quality of the target image meets requirements, improving the success rate of image recognition based on the processed image, and enhancing the user experience.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, specifically to an image processing method, a model training method, an apparatus, and a medium. Background Technology

[0002] In recent years, to facilitate user access to product information, merchants and businesses have often placed items in display cases, simultaneously storing and showcasing goods. When a user needs an item from the display case, they can open it and retrieve it. In this scenario, the display case can identify the item being removed and upload the results. This allows other devices or systems, such as servers or cloud platforms, to use the uploaded identification data to determine the user or item removed from the display case for statistical purposes. Summary of the Invention

[0003] This disclosure provides an image processing method, model training method, device, and medium to solve the problem of poor image processing effect of display cases in related technologies.

[0004] In a first aspect, this disclosure provides an image processing method.

[0005] Specifically, image processing methods include:

[0006] Obtain the image to be processed;

[0007] The image to be processed is input into a pre-trained first target image processing model to obtain the first target feature image output by the first target image processing model.

[0008] Subtract the first target feature image from the image to be processed to obtain the target preprocessed image;

[0009] The preprocessed target image is input into a pre-trained second target image processing model to obtain the second target feature image output by the second target image processing model.

[0010] The target image is obtained by adding the first target feature image and the second target feature image.

[0011] In one implementation of this disclosure, obtaining the image to be processed includes:

[0012] Acquire at least one image captured by the image acquisition device in the display case;

[0013] In response to a payment failure result from payment processing based on at least one captured image, an image to be processed is determined from the at least one captured image.

[0014] In one implementation of this disclosure, in response to a payment failure result from payment processing based on at least one captured image, determining the image to be processed from the at least one captured image includes:

[0015] In response to a payment failure result from payment processing based on at least one acquired image, the overexposed and / or underexposed images in the at least one acquired image are identified as images to be processed.

[0016] In one implementation of this disclosure, both the first target image processing model and the second target image processing model are U-Net models.

[0017] Secondly, this disclosure provides a model training method.

[0018] Specifically, the model training method includes:

[0019] Acquire the training output image and perform overexposure or underexposure processing on the training output image to obtain the training input image;

[0020] The training input image is used as the input of the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image. The first image processing model and the second image processing model are trained. The output of the first image processing model is the first feature image, the input of the second image processing model is the preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image.

[0021] In response to the convergence of both the first image processing model and the second image processing model, the first image processing model is determined as the first target image processing model, and the second image processing model is determined as the second target image processing model.

[0022] In one implementation of this disclosure, the training input image is used as the input to a first image processing model, and the training output image is used as the sum of a first feature image and a second feature image. The first image processing model and the second image processing model are then trained, including:

[0023] The training input image is used as the input to the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image, and then based on M... b =|M o -M i | Obtain the error matrix M b M o M is the training output matrix corresponding to the training output image. i This is the training input matrix corresponding to the training input image;

[0024] According to loss=∑K xyz / (255*n) is used to obtain the training loss, where K xyz The error matrix M b The elements with coordinates (x, y, z) in the middle are n = max(x) * max(y) * max(z);

[0025] The first image processing model and the second image processing model are trained based on the training loss.

[0026] In one implementation of this disclosure, the training output image is obtained, including...

[0027] Acquire at least one image captured by the image acquisition device in the display case;

[0028] In response to the result of payment processing based on at least one acquired image being considered a successful payment, a training output image is determined from at least one acquired image.

[0029] In one implementation of this disclosure, in response to a successful payment result based on at least one acquired image, a training output image is determined from the at least one acquired image, including:

[0030] In response to the payment being successful as a result of payment processing based on at least one acquired image, the clear image from the at least one acquired image is determined as the training output image.

[0031] Thirdly, this disclosure provides an image processing apparatus.

[0032] Specifically, the image processing device includes:

[0033] The image acquisition module is configured to acquire the image to be processed.

[0034] The first feature image acquisition module is configured to input the image to be processed into a pre-trained first target image processing model to obtain the first target feature image output by the first target image processing model.

[0035] The target preprocessing image acquisition module is configured to subtract the first target feature image from the image to be processed in order to obtain the target preprocessing image;

[0036] The second feature image acquisition module is configured to input the target preprocessed image into a pre-trained second target image processing model to obtain the second target feature image output by the second target image processing model.

[0037] The target image acquisition module is configured to add the first target feature image and the second target feature image to obtain the target image.

[0038] Fourthly, this disclosure provides a model training apparatus.

[0039] Specifically, the model training device includes:

[0040] The training image acquisition module is configured to acquire the training output image and perform overexposure or underexposure processing on the training output image to acquire the training input image.

[0041] The model training module is configured to take the training input image as the input of the first image processing model, take the training output image as the sum of the first feature image and the second feature image, and train the first image processing model and the second image processing model. The output of the first image processing model is the first feature image, the input of the second image processing model is the preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image.

[0042] The model determination module is configured to determine the first image processing model as the first target image processing model and the second image processing model as the second target image processing model in response to the convergence of both the first image processing model and the second image processing model.

[0043] Fifthly, this disclosure provides an electronic device including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the method steps as described in any one of the first aspect, any implementation of the first aspect, the second aspect, and any implementation of the second aspect.

[0044] In a sixth aspect, this disclosure provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the method steps as described in any one of the first aspect, any one of the first aspect implementations, the second aspect, and any one of the second aspect implementations.

[0045] In a seventh aspect, this disclosure provides a computer program product including computer instructions that, when executed by a processor, implement the method steps as described in any one of the first aspect, any one of the first aspect implementations, the second aspect, and any one of the second aspect implementations.

[0046] The technical solutions provided in this disclosure may have the following beneficial effects:

[0047] In the technical solution provided in this disclosure, an image to be processed is acquired, and the image to be processed is input into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. The first target feature image is subtracted from the image to be processed to obtain a target preprocessed image. The target preprocessed image is input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target feature image and the second target feature image are added together to obtain the target image. In this scheme, the first target feature image can be understood as an image of the main features that are relatively easy to extract from the image to be processed, while the target preprocessed image can be understood as an image including the regions corresponding to the secondary features that were not extracted from the image to be processed. The second target feature image can be understood as an image of the secondary features, i.e., the details, that are more difficult to extract from the image to be processed. The target image obtained by adding the first target feature image and the second target feature image can be understood as including both the main features that are relatively easy to extract from the image to be processed and the secondary features that are more difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0048] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0049] Other features, objects, and advantages of this disclosure will become more apparent from the following detailed description of non-limiting embodiments, taken in conjunction with the accompanying drawings. In the drawings:

[0050] Figure 1 A schematic structural block diagram of a display cabinet according to an embodiment of the present disclosure is shown.

[0051] Figure 2 A schematic structural block diagram of a motherboard according to an embodiment of the present disclosure is shown.

[0052] Figure 3 A schematic structural block diagram of a control panel according to an embodiment of the present disclosure is shown.

[0053] Figure 4 A schematic structural block diagram of a power management module according to an embodiment of the present disclosure is shown.

[0054] Figure 5 A flowchart of an image processing method according to an embodiment of the present disclosure is shown.

[0055] Figure 6 A flowchart illustrating a model training method according to an embodiment of the present disclosure is shown.

[0056] Figure 7 A schematic structural block diagram of an image processing apparatus according to an embodiment of the present disclosure is shown.

[0057] Figure 8 A schematic structural block diagram of a model training apparatus according to an embodiment of the present disclosure is shown.

[0058] Figure 9 A schematic structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0059] Figure 10 This is a schematic diagram of the structure of a computer system suitable for implementing a method according to an embodiment of the present disclosure. Detailed Implementation

[0060] In the following, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings to enable those skilled in the art to readily implement them. Furthermore, for clarity, portions unrelated to the description of the exemplary embodiments have been omitted from the drawings.

[0061] In this disclosure, it should be understood that terms such as “comprising” or “having” are intended to indicate the presence of features, figures, steps, behaviors, components, parts or combinations thereof disclosed in this specification, and do not preclude the possibility of the presence or addition of one or more other features, figures, steps, behaviors, components, parts or combinations thereof.

[0062] It should also be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0063] As mentioned above, with the development of technology and the improvement of people's living standards, businesses no longer simply place goods on shelves. Instead, to facilitate users' understanding of product information, they can place items in display cases, achieving simultaneous storage and display of goods. When users need to remove items from the display cases, they can open the cases themselves and perform the corresponding operations.

[0064] In recent years, the number of display cases put into operation has gradually increased. During the use of display cases, merchants or enterprises generally need to identify the items in the display cases to determine the items that have been removed from the display cases, and make statistics based on the above information. Based on the statistical results, the quantity and type of the remaining items in the display cases can be known, and further settlement can also be carried out based on the statistical results.

[0065] In one embodiment, after the display case is unlocked, at least one image can be captured by at least one camera installed on the display case, and image recognition can be performed on the image to determine the item removed from the display case based on the image recognition result.

[0066] In related technologies, display cases are often located in environments with varying light intensity (e.g., indoor environments lacking lighting, resulting in prolonged periods of low light; or outdoor environments lacking shading, resulting in prolonged periods of high light). This can lead to poor image quality (e.g., overexposure or underexposure). Therefore, to improve the success rate of image recognition based on images captured by display cases, preprocessing can be performed to ensure that the processed image meets the required quality standards, thereby increasing the success rate of image recognition based on the processed image.

[0067] However, during implementation, the applicant discovered that in the relevant technologies, when processing images, some detailed features in the processed images are often ignored, resulting in these detailed features not being extracted. This reduces the image quality of the processed image, which may not meet the corresponding requirements. Consequently, the success rate of image recognition based on the processed image is reduced, thus harming the user experience.

[0068] In view of the above-mentioned shortcomings, in one embodiment of this disclosure, an image processing method is proposed. The method acquires an image to be processed, inputs the image to be processed into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model, subtracts the first target feature image from the image to be processed to obtain a target preprocessed image, inputs the target preprocessed image into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model, and adds the first target feature image and the second target feature image to obtain the target image. In this scheme, the first target feature image can be understood as an image of the main features that are relatively easy to extract from the image to be processed, while the target preprocessed image can be understood as an image including the regions corresponding to the secondary features that were not extracted from the image to be processed. The second target feature image can be understood as an image of the secondary features, i.e., the details, that are more difficult to extract from the image to be processed. The target image obtained by adding the first target feature image and the second target feature image can be understood as including both the main features that are relatively easy to extract from the image to be processed and the secondary features that are more difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0069] The method provided in this application embodiment can be applied to display cabinets, which can have a temperature control function. The temperature control function can be a cooling function, such as a refrigerated display cabinet, a frozen display cabinet, a refrigerator, a wine cabinet, a cosmetic preservation cabinet, etc.; the temperature control function can also be a heating function, such as a warming cabinet, a heated display cabinet, a hot beverage cabinet, etc. This application embodiment does not limit the specific type of display cabinet.

[0070] For example, Figure 1 A schematic structural block diagram of a display cabinet according to an embodiment of the present disclosure is shown, such as... Figure 1 As shown, the display case 100 may include a compressor 11, a condenser 12, a throttling element 13, and an evaporator 14. The compressor 11, condenser 12, throttling element 13, and evaporator 14 are connected by pipes filled with refrigerant to form a closed pipeline, which constitutes a refrigeration system or heating system capable of circulating refrigerant.

[0071] When the display case is equipped with a refrigeration function, low-temperature, low-pressure gaseous refrigerant flows from the evaporator into the compressor. The compressor compresses the low-temperature, low-pressure gaseous refrigerant, causing the high-temperature, high-pressure gaseous refrigerant to flow into the condenser. The high-temperature, high-pressure gaseous refrigerant exchanges heat with the outside air through the condenser, cooling it into a normal-temperature, high-pressure liquid refrigerant. This liquid refrigerant then flows into a throttling element, which restricts the flow, causing the refrigerant exiting the element to become a low-temperature, low-pressure liquid refrigerant. This low-temperature, low-pressure liquid refrigerant flows into the evaporator, where it exchanges heat with the outside air, evaporating and vaporizing into a low-temperature, low-pressure gaseous refrigerant to absorb heat. In this system, outside air can be introduced into the storage area of ​​the display case through the evaporator, and outside air can be introduced into the outside of the display case through the condenser, thereby transferring the heat in the storage area of ​​the display case to the outside of the display case and cooling the storage area.

[0072] When the display case is equipped with a heating function, low-temperature, low-pressure vaporous refrigerant flows from the condenser into the compressor. The compressor compresses the low-temperature, low-pressure vaporous refrigerant, causing the high-temperature, high-pressure vaporous refrigerant to flow into the evaporator. The high-temperature, high-pressure vaporous refrigerant exchanges heat with the outside air through the evaporator, cooling it into a normal-temperature, high-pressure liquid refrigerant. This liquid refrigerant then flows into a throttling element, which throttles the flow, causing the refrigerant exiting the element to become a low-temperature, low-pressure liquid refrigerant. This low-temperature, low-pressure liquid refrigerant flows into the condenser, where it exchanges heat with the outside air, evaporating and vaporizing into a low-temperature, low-pressure gaseous refrigerant to absorb heat. The system allows outside air to be introduced into the storage area of ​​the display case, and outside air to be introduced into the outside of the display case, thereby transferring heat from outside the display case to the storage area to heat the storage area.

[0073] In one embodiment of this application, the display cabinet includes a cabinet body and a cabinet door, wherein a control board and a power management module may be installed in the cabinet body, and a main board may be installed in the cabinet door.

[0074] In one embodiment of this application, Figure 2 A schematic structural block diagram of a motherboard according to an embodiment of the present disclosure is shown, such as... Figure 2 As shown, the motherboard 200 includes a processor 201, random access memory 202, flash memory 203, wireless LAN Bluetooth module 204, gyroscope 205, pressure sensor 206, microphone 207, speaker 208, camera 209, and cellular communication module 210.

[0075] A processor may include one or more processing units, such as an application processor, a modem processor, a graphics processor, an image signal processor, a controller, a memory, a video codec, a digital signal processor, a baseband processor, and / or a neural network processor. The different processing units may be independent devices or integrated into one or more processors.

[0076] The image signal processor (Image Signal Processor) processes data fed back from the camera. For example, when taking a picture, the shutter is opened, and light is transmitted through the lens to the camera's photosensitive element. The light signal is converted into an electrical signal, which is then transmitted to the Image Signal Processor for processing, transforming it into a visible image. The Image Signal Processor can also perform algorithmic optimizations on image noise, brightness, and skin tone. It can also optimize parameters such as exposure and color temperature of the shooting scene. In some embodiments, the Image Signal Processor can be integrated into the camera itself.

[0077] Digital signal processors (DSPs) are used to process digital signals. Besides digital image signals, they can also process other digital signals. For example, DSPs can be used to perform Fourier transforms on frequency energy.

[0078] Video codecs are used to compress or decompress digital video. A display case can support one or more video codecs. This allows the display case to play or record videos in various encoding formats, such as Moving Picture Experts Group (MPEG) 1, MPEG2, MPEG3, MPEG4, etc.

[0079] Neural network computing processors, by drawing inspiration from the structure of biological neural networks, such as the transmission patterns between neurons in the human brain, can rapidly process input information and continuously learn on their own. These processors can enable applications such as intelligent cognition in display cases, including image recognition, facial recognition, speech recognition, and text understanding.

[0080] In some embodiments, the processor may include one or more interfaces. Interfaces may include integrated circuit interfaces, integrated circuit built-in audio interfaces, pulse code modulation interfaces, universal asynchronous transceiver interfaces, mobile industry processor interfaces, universal input / output interfaces, user identity module interfaces, and / or universal serial bus interfaces, etc.

[0081] Random access memory 202 can be used to store computer executable program code, which includes instructions and data. Processor 201 executes various functional applications and data processing of the display case by running the instructions stored in random access memory 202. Random access memory 202 may include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback, image playback, etc.), etc. The data storage area may store data created during the use of the display case (such as audio data, image data, etc.).

[0082] Flash memory 203 can be used to expand the storage capacity of the display case. Flash memory 203 can communicate with processor 201 via the flash memory interface to achieve data storage functionality. For example, it can store music, video, and other files in the flash memory.

[0083] A minimal system can be constructed using processor 201, random access memory 202, and flash memory 203 to provide the system operating environment.

[0084] The Bluetooth LAN module 204 can provide wireless communication solutions for display cases, including wireless LAN, Bluetooth, GPS, FM, NNHF, and infrared technologies. The Bluetooth LAN module 204 can be one or more devices integrating at least one communication processing module. The Bluetooth LAN module 204 receives electromagnetic waves via an antenna, performs frequency modulation and filtering of the electromagnetic wave signals, and sends the processed signal to the processor 201. The Bluetooth LAN module 204 can also receive signals to be transmitted from the processor 201, perform frequency modulation and amplification, and then convert them into electromagnetic waves for radiation via the antenna. In one embodiment of this application, the Bluetooth LAN module can communicate with a user's terminal.

[0085] The cellular communication module 210 can provide wireless communication solutions, including 2G / 3G / 4G / 5G, for use in display cases. The cellular communication module 210 may include at least one filter, switch, power amplifier, low-noise amplifier, etc. The cellular communication module 210 can receive electromagnetic waves via an antenna, and perform filtering, amplification, and other processing on the received electromagnetic waves before transmitting them to a modem processor for demodulation. The cellular communication module 210 can also amplify the signal modulated by the modem processor and convert it into electromagnetic waves for radiation via the antenna. In some embodiments, at least some functional modules of the cellular communication module 210 may be housed in the processor 201. In some embodiments, at least some functional modules of the cellular communication module 210 and at least some modules of the processor 201 may be housed in the same device. In one embodiment of this application, the cellular communication module 210 can communicate with the cloud server of a display case control service provider.

[0086] Through the Bluetooth wireless LAN module 204 and the cellular communication module 210, the display case can communicate with networks and other devices via wireless communication technology. The wireless communication technology may include Global System for Mobile Communications (GSMA), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband CDMA, Time Division CDMA, and Long Term Evolution (LTE).

[0087] The gyroscope 205 can be used to determine the real-time attitude of the display case doors.

[0088] The pressure sensor 206 is used to sense pressure signals and convert them into electrical signals. In some embodiments, the pressure sensor 206 can be disposed on the display screen. There are many types of pressure sensors 206, such as resistive pressure sensors, inductive pressure sensors, and capacitive pressure sensors. A capacitive pressure sensor may include at least two parallel plates with conductive materials. When a force is applied to the pressure sensor 206, the capacitance between the electrodes changes, and the pressure intensity is determined based on the change in capacitance. When a touch operation is applied to the display screen, the touch operation intensity is detected by the pressure sensor 206, and the touch position can also be calculated based on the detection signal from the pressure sensor 206. In some embodiments, touch operations applied to the same touch position but with different touch operation intensities can correspond to different operation commands. For example, when a touch operation with an intensity less than a first pressure threshold is applied to the beverage selection app icon, a command to view specific beverage information is executed. When a touch operation with an intensity greater than or equal to the first pressure threshold is applied to the beverage selection app icon, a command to purchase a beverage is executed.

[0089] Microphone 207, also known as a "microphone" or "voice transducer," is used to convert sound signals into electrical signals. When making a phone call or sending a voice message, the user can speak by bringing their mouth close to microphone 207, inputting the sound signal into microphone 207. A display case can be equipped with at least one microphone 207. In some embodiments, the display case can be equipped with two microphones 207, which, in addition to collecting sound signals, can also perform noise reduction. In other embodiments, the display case can be equipped with three, four, or more microphones 207, enabling sound signal collection, noise reduction, sound source identification, and directional recording, among other functions. In one embodiment of this application, the microphone 207 can be used to collect the sound of the display case during operation.

[0090] Speaker 208, also known as a "loudspeaker," is used to convert audio electrical signals into sound signals. The display case can play music or announcements via speaker 208.

[0091] Camera 209 is used to capture images, including still images and moving images (i.e., video). An object is projected onto a photosensitive element by an optical image generated through a lens. The photosensitive element can be a charge-coupled device (CCD) or a complementary metal-oxide-semiconductor (CMOS) phototransistor. The photosensitive element converts the light signal into an electrical signal, which is then passed to an image signal processor (ISS) for conversion into a digital image signal. The ISS outputs the digital image signal to a digital signal processor (DSP) for further processing. The DSP converts the digital image signal into image signals in standard formats such as RGB and YUV. In some embodiments, the display case may include one or more cameras 209. In one embodiment of this application, the camera 209 may have a self-heating function to ensure that its lens does not fog up.

[0092] In one embodiment of this application, Figure 3 A schematic structural block diagram of a control panel according to an embodiment of the present disclosure is shown, such as... Figure 3 As shown, the control board 300 includes a power input interface 301, a power output interface 302, a metering chip 303, a microcontroller chip 304, a real-time clock chip, a light switch interface 305, a temperature control switch interface 306, an evaporator fan interface 307, a compressor interface 308, a condenser fan interface 309, a temperature sensor interface 310, a communication interface 311, and a power interface 312.

[0093] The metering chip 303, also known as the power sensor, acquires voltage, current, real-time power, and average power data. A real-time clock chip maintains the time for the microcontroller chip 304. The light switch interface 305 receives control signals from the temperature control cabinet's light switch. The temperature control switch interface 306 receives control signals from the temperature control cabinet's temperature control switch. The evaporator fan interface 307 sends control signals to the evaporator fan to control its operation. The compressor interface 308 sends control signals to the compressor to control its operation. The condenser fan interface 309 sends control signals to the condenser fan to control its operation. The temperature sensor interface 310 receives temperature data from one or more temperature sensors to determine the temperature at one or more locations within the temperature control cabinet.

[0094] In one embodiment of this application, Figure 4 A schematic structural block diagram of a power management module according to an embodiment of the present disclosure is shown, such as... Figure 4As shown, the power management module 400 includes an AC-to-DC conversion module 401, a charging management module 402, and a battery 403. The power management module 400 supplies power to the motherboard and control board and manages the charging and discharging of the battery. The power management module 400 can also monitor parameters such as battery capacity, battery cycle count, and battery health status (leakage current, impedance). In some other embodiments, the power management module 400 may also be located within the processor.

[0095] In one embodiment of this application, the display case further includes a display screen. The display case implements its display function through a graphics processor, a display screen, and an application processor. The graphics processor is a microprocessor for image processing, connected to the display screen and the application processor. The graphics processor is used to perform mathematical and geometric calculations and for graphics rendering. The processor may include one or more graphics processors that execute program instructions to generate or modify display information. The display screen is used to display static images, videos, etc. The display screen includes a display panel. The display panel may be a liquid crystal display, an organic light-emitting diode (OLED), an active matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode, a quantum dot light-emitting diode, etc. In some embodiments, the display case may include one or more displays.

[0096] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the display cabinet. In other embodiments of this application, the display cabinet may include more or fewer components than illustrated, or combine some components, or separate some components, or arrange different components. The illustrated components may be implemented in hardware, software, or a combination of software and hardware. For example, by combining different components, the display cabinet in the embodiments of this application may be any of the following: a retail cabinet, a heated cabinet, a refrigerated cabinet, a freezer, a combination cabinet, or a display case.

[0097] Figure 5 The flowchart illustrates an image processing method according to an embodiment of the present disclosure. This image processing method can be applied to a display case, or to a corresponding server or cloud platform. The server and cloud platform can be understood as being set up by the operator of the display case, and the server or cloud platform can communicate with the display case. It should be noted that the present disclosure does not specifically limit the entity executing the image processing method.

[0098] like Figure 5 As shown, the image processing method includes the following steps S101-S105:

[0099] In step S101, the image to be processed is acquired;

[0100] In one embodiment of this disclosure, the image to be processed can be understood as an image captured by an image acquisition device installed on a display case, or it can be understood as an image captured by other image acquisition devices. Acquiring the image to be processed can be understood as reading a pre-stored image to be processed, or it can be understood as receiving an image to be processed sent by other devices or systems. It should be noted that the image to be processed can be a single image or multiple images.

[0101] In step S102, the image to be processed is input into the pre-trained first target image processing model to obtain the first target feature image output by the first target image processing model;

[0102] In one embodiment of this disclosure, the first target image processing model may be pre-stored in a display case or obtained from other devices or systems. The first target image processing model may be a neural network (NN) model, a convolutional neural network (CNN) model, or a long short-term memory (LSTM) model, etc.

[0103] In step S103, the first target feature image is subtracted from the image to be processed to obtain the target preprocessed image;

[0104] In one embodiment of this disclosure, subtracting the first target feature image from the image to be processed can be understood as performing calculations based on a pre-acquired image subtraction algorithm, substituting the image to be processed and the first target feature image into the algorithm. Alternatively, it can involve inputting the image to be processed and the first target feature image as input to a pre-trained image subtraction model to obtain a target pre-processed image. For example, the target pre-processed image can be obtained by subtracting the pixel value of the corresponding pixel in the first target feature image from the pixel value of the corresponding pixel in the image to be processed.

[0105] In step S104, the preprocessed target image is input into the pre-trained second target image processing model to obtain the second target feature image output by the second target image processing model;

[0106] In one embodiment of this disclosure, the second target image processing model may be pre-stored in a display case or obtained from other devices or systems. The second target image processing model may be a neural network model, a convolutional neural network model, or a long short-term memory network model, etc.

[0107] In step S105, the first target feature image and the second target feature image are added together to obtain the target image.

[0108] In one embodiment of this disclosure, adding the first target feature image and the second target feature image can be understood as performing calculations based on a pre-acquired image addition algorithm, by substituting the first and second target feature images into the algorithm. Alternatively, it can involve using the first and second target feature images as input to a pre-trained image addition model to obtain the target image. For example, the target image can be obtained by adding the pixel values ​​of corresponding pixels in the first target feature image to the pixel values ​​of corresponding pixels in the second target feature image.

[0109] For example, consider a display case used to provide unmanned vending services. When a customer wants to purchase an item from the display case, they can use a mobile communication terminal to scan the QR code on the surface of the display case to access the corresponding cloud authorization server and send a user authorization request to the cloud authorization server, i.e., requesting to purchase the item from the display case. The cloud authorization server then sends user authorization information to the display case based on the user authorization request. The display case receives the user authorization information from the cloud server and unlocks its door in response. After the door is unlocked, the customer can open the door and take the item from the display case through the item entrance / exit. From the time the door is opened until it is closed, the image acquisition device at the display case continuously captures images of the item entrance / exit and the display area to obtain at least one image to be processed. The display case then inputs the image to be processed into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. The first target feature image is subtracted from the image to be processed to obtain a target preprocessed image. The target preprocessed image is then input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target feature image and the second target feature image are added together to obtain the target image. Item recognition is then performed based on the target image to obtain the item location information of at least one item in at least two target images. Based on the item location information, the item that the user wants to buy is determined, and instruction information for the user to pick up the item is uploaded to the cloud. The cloud then performs payment for the purchased item based on the user's instruction information and the previously obtained user authorization request.

[0110] For example, let's illustrate this using a display case to provide unmanned vending services. When maintenance personnel, such as convenience store clerks, need to replenish the display case, they can use a mobile communication terminal to scan the QR code on the surface of the display case to access the corresponding cloud authorization server and send a replenishment request to the server. The cloud authorization server, after verifying the request, sends maintenance personnel authorization information to the display case, allowing them to replenish the items. The display case receives the authorization information from the cloud server and unlocks its door in response. Once unlocked, the maintenance personnel can open the door and move the scanned items into the display case through the item inlet / outlet. After replenishing, the maintenance personnel can close the door. From the time the door is opened until it is closed, the image acquisition device at the display case continuously captures images of the item inlet / outlet and the display area to obtain at least two images for processing. The display case then inputs the image to be processed into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. The first target feature image is subtracted from the image to be processed to obtain a target preprocessed image. The target preprocessed image is then input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target feature image and the second target feature image are then added to obtain a target image. Item recognition is then performed based on the target image to obtain the item location information of at least one item in at least two target images. Based on the item location information, the system determines whether the maintenance personnel have moved the item into or removed the display case from the display case. Maintenance item instruction information for indicating the item is then uploaded to the cloud so that the cloud can update the information of the remaining items in the display case based on the maintenance item instruction information.

[0111] For example, consider a display case used to provide item retrieval services to a target user, where the target user can be understood as a user belonging to a specific unit or department. When a user needs to retrieve an item from the display case, they can use a mobile communication terminal to scan the QR code on the surface of the display case to access the server corresponding to the display case and send an identity authentication request to the server, i.e., requesting authentication of the user's identity; alternatively, they can use a corresponding identification (such as an employee badge, ID card, etc.) to scan the QR code on the display case, which will then send an identity authentication request to the corresponding server. When the display case receives the item retrieval authorization information returned by the server, it can be understood that the server has confirmed the user as the target user, and the display case can unlock its door in response to the item retrieval authorization information. After the door is unlocked, the user can open the door and retrieve the item from the display case through the item access point. After retrieving the item, the user can close the door. From the time the door is opened until it is closed, the image acquisition device on the display case can continuously acquire images of the item access point and the display area to obtain at least two images to be processed. The display case then inputs the image to be processed into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. The first target feature image is subtracted from the image to be processed to obtain a target pre-processed image. The target pre-processed image is then input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target feature image and the second target feature image are added together to obtain the target image. Then, based on the target image, item recognition is performed to obtain the item location information of at least one item in at least two target images. Based on the item location information, the item that the user wants to take is determined, and the user's item taking instruction information is uploaded to the cloud to indicate that the item is taken. The cloud records the item taken by the user based on the user's item taking instruction information and the previously obtained user authorization request.

[0112] In the technical solution provided in this disclosure, an image to be processed is acquired, and the image to be processed is input into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. The first target feature image is subtracted from the image to be processed to obtain a target preprocessed image. The target preprocessed image is input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target feature image and the second target feature image are added together to obtain the target image. In this scheme, the first target feature image can be understood as an image of the main features that are relatively easy to extract from the image to be processed, while the target preprocessed image can be understood as an image including the regions corresponding to the secondary features that were not extracted from the image to be processed. The second target feature image can be understood as an image of the secondary features, i.e., the details, that are more difficult to extract from the image to be processed. The target image obtained by adding the first target feature image and the second target feature image can be understood as including both the main features that are relatively easy to extract from the image to be processed and the secondary features that are more difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0113] In one implementation of this disclosure, obtaining the image to be processed includes:

[0114] Acquire at least one image captured by the image acquisition device in the display case;

[0115] In response to a payment failure result from payment processing based on at least one captured image, an image to be processed is determined from the at least one captured image.

[0116] In one embodiment of this disclosure, determining whether the result of payment processing based on at least one captured image is a payment failure can be understood as obtaining a payment log and obtaining the image identifier used when executing a payment processing procedure that results in a payment failure, and determining the image to be processed from at least one captured image.

[0117] In the technical solution provided in this disclosure, considering that when the payment processing based on at least one acquired image results in payment failure, it is because the image quality of some or all of the images in the at least one acquired image is low, in order to improve the image quality of some or all of the images in the at least one acquired image, by determining the image to be processed in the at least one acquired image, it can be ensured that the image quality of the processed target image is better and can meet the corresponding requirements, which helps to improve the success rate of image recognition, and thus helps to improve the payment success rate.

[0118] In one implementation of this disclosure, in response to a payment failure result from payment processing based on at least one captured image, determining the image to be processed from the at least one captured image includes:

[0119] In response to a payment failure result from payment processing based on at least one acquired image, the overexposed and / or underexposed images in the at least one acquired image are identified as images to be processed.

[0120] In one embodiment of this disclosure, at least one overexposed image in the captured images can be understood as an image with excessive brightness and a washed-out display effect caused by the image capturing device using an excessively large aperture or a excessively slow shutter speed during image capture. At least one underexposed image in the captured images can be understood as an image with insufficient brightness and a blackish display effect caused by the image capturing device using an excessively small aperture or a excessively fast shutter speed during image capture.

[0121] At least one overexposed and / or underexposed image in the acquired image can be understood as being determined based on a pre-acquired corresponding algorithm, or it can be understood as being determined based on a pre-trained corresponding model.

[0122] In the technical solution provided in this disclosure, considering that some images in at least one acquired image may still have good image quality and do not require image processing, by responding to the result of payment processing based on at least one acquired image as payment failure, the overexposed and / or underexposed images in at least one acquired image are identified as images to be processed, which can minimize the number of images to be processed and reduce the computational cost.

[0123] In one implementation of this disclosure, both the first target image processing model and the second target image processing model are U-Net (Convolutional Networks for Biomedical Image Segmentation) models.

[0124] Figure 6 The flowchart illustrates a model training method according to an embodiment of this disclosure. This model training method can be applied to a display case, or to a corresponding server or cloud. The server and cloud can be understood as being set up by the operator of the display case, and the server or cloud can communicate with the display case. It should be noted that this disclosure does not specifically limit the entity executing the image training method.

[0125] like Figure 6 As shown, the image processing method includes the following steps S201-S203:

[0126] In step S201, the training output image is obtained, and the training output image is overexposed or underexposed to obtain the training input image.

[0127] In one embodiment of this disclosure, the training output image can be understood as a relatively clear image without overexposure or underexposure. The training output image can be understood as an image captured by an image acquisition device mounted on a display case, or it can be understood as an image captured by other image acquisition devices. Acquiring the training output image can be understood as reading a pre-stored training output image, or it can be understood as receiving a training output image sent by other devices or systems. It should be noted that the training output image can be a single image or multiple images.

[0128] Overexposure processing of the training output image can be understood as either substituting the training output image into a pre-acquired overexposure processing algorithm for calculation, or as using the training output image as input to a pre-trained overexposure processing model to obtain the training input image.

[0129] Underexposing the training output image can be understood as performing calculations based on a pre-acquired underexposure processing algorithm, using the training output image as input, or as using the training output image as input to a pre-trained underexposure processing model to obtain the training input image.

[0130] In step S202, the training input image is used as the input of the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image to train the first image processing model and the second image processing model.

[0131] The output of the first image processing model is the first feature image, the input of the second image processing model is the preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image.

[0132] In one embodiment of this disclosure, the first image processing model may be pre-stored in a display case or obtained from other devices or systems. The first image processing model may be a neural network model, a convolutional neural network model, or a long short-term memory network model, etc.

[0133] Similarly, the second image processing model can be pre-stored in a display case or obtained from other devices or systems. The second image processing model can be a neural network model, a convolutional neural network model, or a long short-term memory network model, etc.

[0134] In one embodiment of this disclosure, the sum of the first feature image and the second feature image can be understood as the pixel value of the corresponding pixel in the training output image, which is the sum of the pixel value of the corresponding pixel in the first feature image and the pixel value of the corresponding pixel in the second feature image.

[0135] In one embodiment of this disclosure, subtracting the first feature image from the training input image can be understood as performing calculations based on a pre-acquired image addition algorithm, substituting the training input image and the first feature image into the calculation. Alternatively, it can involve using the training input image and the first feature image as input to a pre-trained image addition model to obtain the target image. For example, a preprocessed image can be obtained by subtracting the pixel value of the corresponding pixel in the first feature image from the pixel value of the corresponding pixel in the training input image.

[0136] In step S203, in response to the convergence of both the first image processing model and the second image processing model, the first image processing model is determined as the first target image processing model, and the second image processing model is determined as the second target image processing model.

[0137] In the technical solution provided in this disclosure, a training input image is obtained by acquiring a training output image and performing overexposure or underexposure processing on the training output image. This training input image is then used as the input to a first image processing model, and the training output image is used as the sum of a first feature image and a second feature image to train both the first and second image processing models. Upon convergence of both the first and second image processing models, the first image processing model is determined as the first target image processing model, and the second image processing model is determined as the second target image processing model. This ensures that when image processing is performed based on the trained first and second target image processing models, the first target image processing model... The model can extract a first target feature image from the image to be processed, including the main features that are relatively easy to extract. The first target image processing model can extract a second target feature image from the target preprocessed image obtained by subtracting the first target feature image from the image to be processed, including the secondary features, i.e., the details, that were not extracted in the image to be processed. This allows the target image to be obtained by adding the first target feature image and the second target feature image, which includes both the main features that are relatively easy to extract from the image to be processed and the secondary features that are relatively difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0138] In one implementation of this disclosure, the training input image is used as the input to the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image. The first image processing model and the second image processing model are trained through the following steps:

[0139] The training input image is used as the input to the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image, and then based on M... b =|M o -M i | Obtain the error matrix M b M o M is the training output matrix corresponding to the training output image. i This is the training input matrix corresponding to the training input image;

[0140] According to loss=∑K xyz / (255*n) is used to obtain the training loss, where K xyz The error matrix M b The elements with coordinates (x, y, z) in the middle are n = max(x) * max(y) * max(z);

[0141] The first image processing model and the second image processing model are trained based on the training loss.

[0142] In the technical solution provided in this disclosure, by using the training input image as the input of the first image processing model, using the training output image as the sum of the first feature image and the second feature image, and obtaining the error matrix, obtaining the training loss based on the error matrix, and training the first image processing model and the second image processing model according to the training loss, the efficiency of training the first image processing model and the second image processing model can be improved.

[0143] In one implementation of this disclosure, the training output image can be obtained through the following steps.

[0144] Acquire at least one image captured by the image acquisition device in the display case;

[0145] In response to the result of payment processing based on at least one acquired image being considered a successful payment, a training output image is determined from at least one acquired image.

[0146] In one embodiment of this disclosure, determining whether the result of payment processing based on at least one acquired image is a successful payment can be understood as obtaining payment logs and obtaining the image identifier used when executing the payment processing process with a successful payment result, and determining the training output image from at least one acquired image.

[0147] In the technical solution provided in this disclosure, considering that the image quality of the at least one acquired image is high when the payment processing based on the at least one acquired image results in a successful payment, the training output image can be determined from the at least one acquired image in response to the successful payment processing based on the at least one acquired image. This ensures that the image quality of the training output image is good and can meet the corresponding requirements, thereby helping to improve training efficiency.

[0148] In one implementation of this disclosure, in response to the result of payment processing based on at least one acquired image indicating successful payment, determining the training output image from the at least one acquired image can be achieved through the following steps:

[0149] In response to the payment being successful as a result of payment processing based on at least one acquired image, the clear image from the at least one acquired image is determined as the training output image.

[0150] In one embodiment of this disclosure, determining a clear image in at least one acquired image can be understood as follows: based on a pre-acquired clear image algorithm, at least one acquired image is substituted into the algorithm for calculation to obtain corresponding clear image indication information, and the clear image in at least one acquired image is determined based on the clear image indication information; alternatively, at least one acquired image can be used as input to a pre-trained clear image model to obtain corresponding clear image indication information, and the clear image in at least one acquired image is determined based on the clear image indication information.

[0151] In the technical solution provided in this disclosure, by responding to the result of payment processing based on at least one acquired image as a successful payment, and determining the clear image in at least one acquired image as the training output image, it is possible to further ensure that the image quality of the training output image is good and can meet the corresponding requirements, which helps to improve training efficiency.

[0152] The following are embodiments of the apparatus disclosed herein, which can be used to execute embodiments of the method disclosed herein.

[0153] Figure 7 A schematic structural block diagram of an image processing apparatus according to an embodiment of the present disclosure is shown. This image processing apparatus can be implemented as part or all of an electronic device through software, hardware, or a combination of both. Figure 7 As shown, the image processing apparatus includes:

[0154] The image acquisition module 301 is configured to acquire the image to be processed.

[0155] The first feature image acquisition module 302 is configured to input the image to be processed into a pre-trained first target image processing model to obtain the first target feature image output by the first target image processing model.

[0156] The target preprocessing image acquisition module 303 is configured to subtract the first target feature image from the image to be processed in order to obtain the target preprocessing image;

[0157] The second feature image acquisition module 304 is configured to input the target preprocessed image into a pre-trained second target image processing model to obtain the second target feature image output by the second target image processing model.

[0158] The target image acquisition module 305 is configured to add the first target feature image and the second target feature image to obtain the target image.

[0159] The above technical solution involves acquiring an image to be processed, inputting the image to be processed into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model, subtracting the first target feature image from the image to be processed to obtain a target preprocessed image, inputting the target preprocessed image into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model, and adding the first target feature image and the second target feature image to obtain the target image. In this scheme, the first target feature image can be understood as an image of the main features that are relatively easy to extract from the image to be processed, while the target preprocessed image can be understood as an image including the regions corresponding to the secondary features that were not extracted from the image to be processed. The second target feature image can be understood as an image of the secondary features, i.e., the details, that are more difficult to extract from the image to be processed. The target image obtained by adding the first target feature image and the second target feature image can be understood as including both the main features that are relatively easy to extract from the image to be processed and the secondary features that are more difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0160] Figure 8 A schematic structural block diagram of a model training apparatus according to an embodiment of the present disclosure is shown. This model training apparatus can be implemented as part or all of an electronic device through software, hardware, or a combination of both. Figure 8 As shown, the model training device includes:

[0161] The training image acquisition module 401 is configured to acquire the training output image and perform overexposure or underexposure processing on the training output image to acquire the training input image.

[0162] The model training module 402 is configured to use the training input image as the input of the first image processing model and the training output image as the sum of the first feature image and the second feature image to train the first image processing model and the second image processing model. The output of the first image processing model is the first feature image, the input of the second image processing model is the preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image.

[0163] The model determination module 403 is configured to determine the first image processing model as the first target image processing model and the second image processing model as the second target image processing model in response to the convergence of both the first image processing model and the second image processing model.

[0164] In the technical solution provided in this disclosure, a training input image is obtained by acquiring a training output image and performing overexposure or underexposure processing on the training output image. This training input image is then used as the input to a first image processing model, and the training output image is used as the sum of a first feature image and a second feature image to train both the first and second image processing models. Upon convergence of both the first and second image processing models, the first image processing model is determined as the first target image processing model, and the second image processing model is determined as the second target image processing model. This ensures that when image processing is performed based on the trained first and second target image processing models, the first target image processing model... The model can extract a first target feature image from the image to be processed, including the main features that are relatively easy to extract. The first target image processing model can extract a second target feature image from the target preprocessed image obtained by subtracting the first target feature image from the image to be processed, including the secondary features, i.e., the details, that were not extracted in the image to be processed. This allows the target image to be obtained by adding the first target feature image and the second target feature image, which includes both the main features that are relatively easy to extract from the image to be processed and the secondary features that are relatively difficult to extract from the image to be processed. This ensures that the image quality of the target image meets the corresponding requirements, thereby improving the success rate of image recognition based on the processed image and improving the user experience.

[0165] This disclosure also discloses an electronic device. Figure 9 A schematic structural block diagram of an electronic device according to an embodiment of the present disclosure is shown, such as... Figure 9 As shown, the electronic device includes a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the above method steps.

[0166] Figure 10 This is a schematic diagram of the structure of a computer system suitable for implementing a method according to an embodiment of the present disclosure. For example... Figure 10 As shown, the computer system includes a processing unit that can execute various processes described above based on a program stored in a read-only memory (ROM) or a program loaded from a storage portion into a random access memory (RAM). The RAM also stores various programs and data required for the operation of the computer system. The processing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0167] The following components are connected to the I / O interface: input sections including keyboards, mice, etc.; output sections including cathode ray tubes (CRTs), liquid crystal displays (LCDs), and speakers; storage sections including hard disks; and communication sections including network interface cards such as LAN cards and modems. The communication sections perform communication processing via networks such as the Internet. Drives are also connected to the I / O interface as needed. Removable media, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive as needed so that computer programs read from them can be installed into the storage section as needed. The processing unit can be implemented as a CPU, GPU, TPU, FPGA, NPU, etc.

[0168] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0169] The units or modules described in the embodiments of this disclosure can be implemented in software or hardware. The described units or modules can also be located in a processor, and the names of these units or modules do not necessarily constitute a limitation on the unit or module itself.

[0170] In another aspect, this disclosure also provides a computer-readable storage medium, which may be a computer-readable storage medium included in the apparatus described in the above embodiments; or it may be a standalone computer-readable storage medium not assembled into a device. The computer-readable storage medium stores one or more programs that are used by one or more processors to perform the methods described in this disclosure.

[0171] In addition, this disclosure also provides a computer program product storing a computer program that, when executed by a processor, enables the processor to at least implement the methods provided in the foregoing embodiments.

[0172] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

Claims

1. An image processing method, characterized in that, The method includes: Obtain the image to be processed; The image to be processed is input into a pre-trained first target image processing model to obtain a first target feature image output by the first target image processing model. Subtract the first target feature image from the image to be processed to obtain the target preprocessed image; The preprocessed target image is input into a pre-trained second target image processing model to obtain a second target feature image output by the second target image processing model. The first target image processing model and the second target image processing model are trained through the following steps: Acquire the training output image and perform overexposure or underexposure processing on the training output image to obtain the training input image; The training input image is used as the input of the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image to train the first image processing model and the second image processing model. The output of the first image processing model is the first feature image, the input of the second image processing model is the preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image. The step of using the training input image as input to the first image processing model and the training output image as the sum of the first feature image and the second feature image to train the first image processing model and the second image processing model includes: The training input image is used as the input to the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image, and according to M... b =|M o -M i | Obtain the error matrix M b M o M is the training output matrix corresponding to the training output image. i This is the training input matrix corresponding to the training input image; According to loss=ΣK xyz / (255 n) Obtain the training loss, where K xyz The error matrix M b The elements with coordinates (x, y, z) in the middle, n = max(x). max(y) max(z); The first image processing model and the second image processing model are trained based on the training loss. In response to the convergence of both the first image processing model and the second image processing model, the first image processing model is determined as the first target image processing model, and the second image processing model is determined as the second target image processing model. The first target feature image is added to the second target feature image to obtain the target image.

2. The image processing method according to claim 1, characterized in that, The process of acquiring the image to be processed includes: Acquire at least one image captured by the image acquisition device in the display case; In response to a payment failure result from payment processing based on the at least one captured image, the image to be processed is determined from the at least one captured image.

3. The image processing method according to claim 2, characterized in that, In response to a payment failure result from payment processing based on the at least one captured image, determining the image to be processed from the at least one captured image includes: In response to a payment failure result from payment processing based on the at least one acquired image, the overexposed and / or underexposed images in the at least one acquired image are identified as the images to be processed.

4. The image processing method according to claim 1, characterized in that, Both the first target image processing model and the second target image processing model are U Net model.

5. The image processing method according to claim 1, characterized in that, The process of obtaining the training output image includes: Acquire at least one image captured by the image acquisition device in the display case; In response to the result of payment processing based on the at least one acquired image being successful, the training output image is determined from the at least one acquired image.

6. The image processing method according to claim 5, characterized in that, The step of determining the training output image from the at least one acquired image in response to a successful payment processing result based on the at least one acquired image includes: In response to the payment being successful as a result of payment processing based on the at least one acquired image, the clear image in the at least one acquired image is determined as the training output image.

7. An image processing apparatus, characterized in that, include: The image acquisition module is configured to acquire the image to be processed. The first feature image acquisition module is configured to input the image to be processed into a pre-trained first target image processing model to obtain the first target feature image output by the first target image processing model. The target preprocessed image acquisition module is configured to subtract the first target feature image from the image to be processed to obtain the target preprocessed image; The second feature image acquisition module is configured to input the target preprocessed image into a pre-trained second target image processing model to obtain the second target feature image output by the second target image processing model. The target image acquisition module is configured to add the first target feature image and the second target feature image to obtain the target image; The image processing device further includes: The training image acquisition module is configured to acquire the training output image and perform overexposure or underexposure processing on the training output image to acquire the training input image. The model training module is configured to use the training input image as the input of a first image processing model, and the training output image as the sum of a first feature image and a second feature image to train the first image processing model and the second image processing model. The output of the first image processing model is the first feature image, the input of the second image processing model is a preprocessed image obtained by subtracting the first feature image from the training input image, and the output of the second image processing model is the second feature image. The model determination module is configured to determine the first image processing model as the first target image processing model and the second image processing model as the second target image processing model in response to the convergence of both the first image processing model and the second image processing model. When the model training module uses the training input image as input to the first image processing model and the training output image as the sum of the first feature image and the second feature image to train the first image processing model and the second image processing model, the model training module is specifically configured as follows: The training input image is used as the input of the first image processing model, and the training output image is used as the sum of the first feature image and the second feature image. The error matrix Mb is obtained according to Mb=|Mo-Mi|, where Mo is the training output matrix corresponding to the training output image and Mi is the training input matrix corresponding to the training input image. According to loss=ΣKxyz / (255) n) Obtain the training loss, where Kxyz are the elements with coordinates (x, y, z) in the error matrix Mb, and n = max(x). max(y) max(z); The first image processing model and the second image processing model are trained based on the training loss.

8. An electronic device, characterized in that, It includes a memory and at least one processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the at least one processor to implement claim 1.

6. The method steps described in any one of the following methods.

9. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instruction is executed by the processor, it implements claim 1.

6. The method steps described in any one of the following methods.

Images