Image acquisition method and image processing method based on dual-channel imaging device

Abstract

The present invention discloses an image acquisition method and an image processing method based on a dual-channel imaging device. The image acquisition method utilizes the characteristic that the original image pairs acquired by the dual-channel imaging device have target reciprocity, that is, the blurred-clear target in one image just corresponds one-to-one to the clear-fuzzy target in the other image in the image pair, and acquires original image pairs at different focal plane distances through the dual-channel imaging device. Then, the image quality is further enhanced and the blurred-clear ROI image pairs are extracted through the image processing method to establish a blurred-clear image pair dataset with a defocus distance label. The present invention only needs to step-by-step adjust the focal plane of one imaging system of the dual-channel imaging device to n positions to obtain image pairs at 2n defocus distances, and simply, conveniently and efficiently establish clear-fuzzy image pairs with different defocus distances, which solves the problem of difficulty in obtaining large-scale high-quality datasets and high dataset acquisition costs.

Description

Technical Field

[0001] The present invention belongs to the field of image processing technology, and more specifically, relates to an image acquisition method and an image processing method based on a dual-channel imaging device. Background Art

[0002] Expanding the imaging depth of field without changing hardware requirements will greatly improve observation efficiency in applications requiring high depth of field. Methods for extending the depth of field of imaging systems using image restoration computational equivalents can be broadly categorized into those based on physical modeling and those driven by data-driven deep learning.

[0003] Data-driven depth of field extension methods use large-scale data to train DNN models. Theoretically, they can directly achieve end-to-end clear restoration of images at a certain defocus distance, effectively extending the depth of field. Training these models requires a large amount of high-quality training data.

[0004] Related technologies typically use expensive, precision automated microscopes to collect image stacks at varying degrees of defocus from a fixed target to create microscopic image datasets, or rely on specialized dual-pixel light field cameras to capture natural scene image datasets with embedded defocus distance information. However, these training data acquisition methods are very time- and financially expensive. Furthermore, in order for the trained models to achieve good generalization performance in real-world observations, the datasets must be substantial in number and category. Consequently, existing technologies face challenges in acquiring large-scale, high-quality datasets, as well as high costs associated with acquiring them. Summary of the Invention

[0005] In response to the shortcomings of the related art, the present invention provides an image acquisition method and an image processing method based on a dual-channel imaging device, aiming to solve the problems existing in the related art of difficulty in obtaining large-scale high-quality data sets and high data set acquisition costs.

[0006] The technical solution is as follows:

[0007] According to one aspect of the present application, an image acquisition method based on a dual-channel imaging device, the dual-channel imaging device including two imaging systems, the method including the following steps: adjusting the focal plane positions of the two imaging systems so that the focal plane spacing of the two imaging systems is Δz; exposing the two imaging systems synchronously for imaging to obtain original image pairs when the focal plane spacing is Δz; and adjusting the focal plane positions of the two imaging systems so that the two imaging systems synchronously for imaging to obtain original image pairs I-I' at different focal plane spacings.

[0008] According to one aspect of the present application, an image processing method based on a dual-channel imaging device includes a two-channel imaging system. The method includes the following steps: obtaining an original image pair II' at different focal plane distances; the original image pair II' includes a first channel image I and a second channel image I'; registering the first channel image I with the second channel image I' to obtain a registered image pair I R -I'; for the registered image pair I R -I' is preprocessed to obtain the processed image pair I P -I P '; The image pair I P -I P 'Including the first processed image I P and the second processed image I P '; for the processed image pair I P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image; the first processed image I P The ROI image and the second processed image I P ' ROI images are evaluated for clarity, thereby respectively screening out the first coordinate and the second coordinate of the clear ROI image; using the first coordinate of the clear ROI image to respectively select the first processed image I P and the second processed image I P ', and use the second coordinates of the clear ROI image to respectively extract the first processed image I P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

[0009] According to one aspect of the present application, an image processing system based on a dual-channel imaging device includes a two-channel imaging system, and the image processing system includes: an acquisition module for obtaining an original image pair I-I' at different focal plane distances; the original image pair I-I' includes a first channel image I and a second channel image I'; a registration module for registering the first channel image I with the second channel image I' to obtain a registered image pair I R -I'; pre-processing module, used for the registration image I R -I' is preprocessed to obtain the processed image pair I P -I P '; The image pair I P -I P 'Including the first processed image IP and the second processed image I P '; Target detection module, for the processed image I P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image; evaluation module for the first processing image I P The ROI image and the second processed image I P ' ROI images are respectively evaluated for clarity, thereby respectively screening out the first coordinate and the second coordinate of the clear ROI image; a cropping module is used to use the first coordinate of the clear ROI image to respectively select the first processed image I P and the second processed image I P ', and use the second coordinates of the clear ROI image to respectively extract the first processed image I P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

[0010] According to one aspect of the present application, an image processing device includes: at least one processor, at least one memory, and at least one communication bus, wherein a computer program is stored on the memory, and the processor reads the computer program in the memory through the communication bus; when the computer program is executed by the processor, the image processing method described above is implemented.

[0011] According to one aspect of the present application, a storage medium stores a computer program thereon, wherein the computer program implements the image processing method described above when executed by a processor.

[0012] According to one aspect of the present application, a computer program product includes a computer program, the computer program is stored in a storage medium, a processor of a computer device reads the computer program from the storage medium, and the processor executes the computer program, so that the computer device implements the image processing method described above when executing the computer program.

[0013] The present invention has the following beneficial effects:

[0014] 1. Traditional image acquisition methods can only capture raw image pairs with different focal plane distances for static targets, but cannot capture raw image pairs with different focal plane distances for moving and changing targets. In addition, the imaging system needs to be frequently adjusted during the image acquisition process, which is cumbersome and inefficient. The present invention provides an image acquisition method based on a dual-channel imaging device, which can capture raw image pairs with different focal plane distances not only for static targets but also for moving or changing targets. In the laboratory, it is possible to simulate the in-situ imaging conditions of the ocean to efficiently capture a large number of dual-channel raw image pairs with different focal plane distances. It only needs to step-by-step adjust the focal plane of one channel of the dual-channel imaging device to n positions to achieve the acquisition of raw image pairs at n focal plane distances. This method is simple and easy to operate, has high image acquisition efficiency, and greatly reduces the difficulty of obtaining real data sets.

[0015] 2. The present invention provides an image processing method based on a dual-channel imaging device, which cleverly utilizes the target reciprocity between the original image pairs obtained by the dual-channel imaging device, that is, the blurred-clear target in one image corresponds exactly to the clear-blurred target in the other image in the image pair. It simply, conveniently and efficiently establishes clear-blurred image pairs with different defocus distances, and all the defocused images are labeled with accurate defocus distances, which solves the problems in the prior art of lack of model training data sets, difficulty in obtaining large-scale high-quality data sets, and high data set acquisition costs. It is simple and easy to implement, with low time and economic costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of an implementation environment involved in an image acquisition method based on a dual-channel imaging device provided in an embodiment of the present application;

[0017] Figure 2 This is a flow chart of an image acquisition method based on a dual-channel imaging device provided in an embodiment of the present application;

[0018] Figure 3 A pair of original plankton images collected using an image acquisition method based on a dual-channel imaging device provided in an embodiment of the present application;

[0019] Figure 4 This is a flow chart of an image processing method based on a dual-channel imaging device provided in an embodiment of the present application;

[0020] Figure 5 It is aimed at Figure 4 A schematic diagram of an example of establishing a label with a defocus distance for the image obtained in step 400;

[0021] Figure 6 An image processing method based on a dual-channel imaging device provided in an embodiment of the present application is used to Figure 3The defocused blur and focused clear image pair of the plankton target is obtained by processing the original image pair of the plankton;

[0022] Figure 7 is a structural diagram of an image processing system provided in an embodiment of the present application;

[0023] Figure 8 This is a structural diagram of an image processing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0025] See also Figure 1 , which shows a schematic diagram of an implementation environment involved in an image acquisition method based on a dual-channel imaging device provided in an embodiment of the present application. The implementation environment includes a two-channel imaging system 320 and a beam splitter 300.

[0026] The light adjusted by the imaging target is split into two beams by the beam splitter 300 and enters the two imaging systems 320 respectively.

[0027] In this implementation environment, the focal plane positions of the two imaging systems 320 are adjusted so that the focal plane spacing of the two imaging systems 320 is a specific value; the two imaging systems 320 are synchronously exposed and imaged to obtain original image pairs with a specific focal plane spacing; the focal plane positions of the two imaging systems 320 are adjusted so that the two imaging systems 320 are synchronously exposed and imaged to obtain original image pairs I-I' with different focal plane spacings.

[0028] See also Figure 2 , which shows a flow chart of an image acquisition method based on a dual-channel imaging device provided by an embodiment of the present application. The dual-channel imaging device includes two imaging systems. The image acquisition method may include the following steps:

[0029] Step 200, adjusting the focal plane positions of the two imaging systems so that the focal plane distance between the two imaging systems is Δz;

[0030] Step 220 , causing the two imaging systems to perform synchronous exposure imaging to obtain an original image pair with a focal plane spacing of Δz;

[0031] Step 240 , adjusting the focal plane positions of the two imaging systems so that the two imaging systems are exposed and imaged synchronously, and obtaining original image pairs II′ at different focal plane distances.

[0032] In a possible implementation, original image pairs II' with focal plane spacings of 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 mm are acquired by the image acquisition method.

[0033] In one possible implementation, the exposure time of the synchronous exposure imaging of the two imaging systems is 400 μs. The 400 μs exposure time can avoid motion blur caused by the motion of the imaging target, thereby obtaining a high-quality original image pair.

[0034] See also Figure 3 , which shows a pair of original plankton images collected using an image acquisition method based on a dual-channel imaging device provided by an embodiment of the present application.

[0035] Traditional image acquisition methods can only capture raw image pairs with different focal plane spacings for static targets, but cannot capture raw image pairs with different focal plane spacings for moving and changing targets. Furthermore, the imaging system needs to be frequently adjusted during image acquisition, which is cumbersome and inefficient. The present invention provides an image acquisition method based on a dual-channel imaging device that can capture raw image pairs with different focal plane spacings not only for static targets but also for moving or changing targets. This method can efficiently capture a large number of dual-channel raw image pair datasets with different focal plane spacings in a laboratory, simulating in-situ ocean imaging conditions. This method is simple and easy to implement, has high image acquisition efficiency, and greatly reduces the difficulty of obtaining real datasets.

[0036] See also Figure 4 , which shows a flow chart of an image processing method based on a dual-channel imaging device provided by an embodiment of the present application. The dual-channel imaging device includes two imaging systems. The image processing method may include the following steps:

[0037] Step 300 : obtaining an original image pair II′ at different focal plane distances; the original image pair II′ includes a first-pass image I and a second-pass image I′.

[0038] The focal plane spacing can be set as needed, for example, the focal plane spacing can be discrete focal plane spacings such as 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 mm, etc. The original image pair is generated by the two imaging systems of the dual-channel device.

[0039] Step 320: align the first path image I with the second path image I' to obtain a registered image pair I R -I'.

[0040] In one possible implementation, image matching is performed using a 3D-printed target. For example, the target is a 11mm*12mm*2mm cube with 1mm*1mm*1mm cubes spaced evenly across one side. The target is imaged with the focal plane spacing between the two imaging systems at zero. Both target images are clearly focused. Point and line features are extracted from the cubes on the target. After manually matching the features, the affine transformation matrix is ​​solved to establish a registration model for the two imaging systems. This registration model is then applied to register original image pairs at other focal plane spacings.

[0041] Step 340, register the image pair I R -I' is preprocessed to obtain the processed image pair I P -I P '; The image pair I P -I P 'Including the first processed image I P and the second processed image I P '.

[0042] The preprocessing may include background subtraction, white balance, etc., which are not limited here.

[0043] Step 360, the processed image is P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image.

[0044] ROI (region of interest) refers to the region of interest, that is, the specific area in the image that needs to be processed. P -I P 'Find the target object. Target detection can be performed using a global threshold method. After finding the target object, extract the target object's region to obtain a target ROI image.

[0045] Step 380: process the first processed image I P The ROI image and the second processed image I P The clarity of the ROI images is evaluated respectively, thereby screening out the first coordinate and the second coordinate of the clear ROI image respectively.

[0046] The ROI image contains clearly focused ROI images and out-of-focus ROI images. The clarity evaluation is to filter out the clear ROI images. The first coordinate refers to the value of the image from the first processed image I PThe coordinates of the clear ROI image filtered out by the ROI image are as follows: P 'The coordinates of the clear ROI image filtered out by the ROI image.

[0047] Step 400: using the first coordinates of the clear ROI image, the first processed image I P and the second processed image I P 'Crop, and use the second coordinates of the clear ROI image to respectively extract the image from the first processing image I P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

[0048] According to the first coordinate, respectively in the first processing image I P and the second processed image I P 'Crop from the first processed image I P Cut out a clear ROI image from the second processed image I P 'Crop out the defocused blurred ROI image and obtain the clear-blurred image pairs at different defocus distances.

[0049] According to the second coordinates, the first processing image I P and the second processed image I P 'Crop and get the first processed image I P The defocused blurred ROI image cut out from the second processed image I P 'The clear ROI image cropped out.

[0050] Through the above embodiments, the characteristic of target reciprocity between the original image pairs obtained by the dual-channel imaging device is cleverly utilized, that is, the blurred-clear target in one image corresponds exactly to the clear-blurred target in the other image in the image pair. Based on the original image pairs at n focal plane intervals, 2n clear-blurred image pairs at defocus distances can be obtained, and a blurred-clear image pair dataset with known defocus distance labels is established. This is simple and easy, with low time and economic costs, and solves the problems in the prior art of difficulty in obtaining large-scale high-quality datasets and high dataset acquisition costs.

[0051] In an exemplary embodiment, step 320 includes:

[0052] Step 321: Use the target to establish a registration model for the two-way imaging system.

[0053] The target may be a 3D-printed cube, for example, a cube of 11mm*12mm*2mm, with small cubes of 1mm*1mm*1mm distributed at equal intervals on one side of the target.

[0054] The target is imaged when the focal plane distance between the two imaging systems is 0. At this time, both images are clearly focused. The point and line features in the two images are extracted by using the cube on the target. After manually matching the features, the affine transformation matrix is ​​solved to establish the registration model of the two imaging systems.

[0055] Step 323: align the first path image I with the second path image I' using the registration model to obtain a registered image pair I R -I'.

[0056] In order to reduce the registration error of the original image pairs under different focal plane distances, only the positions of the two imaging systems in the optical axis direction are changed during the movement of the focal plane.

[0057] In an exemplary embodiment, after step 400, the method further includes:

[0058] Step 401 : Cleaning the clear image of the clear-blurred image pair at different defocus distances, removing the image that does not meet the set requirements in the clear image, and removing the blurred image corresponding to the removed image in the clear image.

[0059] Among them, images that do not meet the set requirements may be images that are still blurry when evaluated by the human eye, images with multiple targets, and images with some targets blurred and some clear, etc., which are not limited here.

[0060] Through the above embodiments, the quality of clear-blurred image pairs at different defocus distances is further improved, and a high-quality image pair dataset can improve the efficiency of model training and the performance of the model.

[0061] In an exemplary embodiment, after step 400, the method further includes:

[0062] Step 403 : Using the defocus distance as a label, a dataset of clear-blurred ROI in-situ image pairs is established.

[0063] The defocus distance includes positive defocus distance and negative defocus distance, and the focal plane distance is the absolute value of the defocus distance.

[0064] In one possible implementation, using the defocus distance as a label can be: when the imaging target is located between the focal plane and the two-way imaging system, the defocus distance is equal to the focal plane distance; when the imaging target is located outside the focal plane, the defocus distance is the opposite of the focal plane distance.

[0065] like Figure 5 As shown, target 1 is located outside the focal plane 2, and its defocus distance is the opposite of the focal plane spacing; target 2 is located between the focal plane 1 and the two-way imaging system, and its defocus distance is equal to the focal plane spacing.

[0066] Through the above embodiment, labels are attached to the acquired clear-blurred image pairs with different defocus distances, so that all defocused images are labeled with accurate defocus distances.

[0067] See also Figure 6 , which shows the use of an image processing method based on a dual-channel imaging device provided by an embodiment of the present application Figure 3 The defocused blurred and focused clear image pairs of the plankton targets are obtained by processing the original plankton images.

[0068] The following are embodiments of the apparatus of the present application, which can be used to perform the image processing method involved in the present application. For details not disclosed in the embodiments of the apparatus of the present application, please refer to the method embodiments of the image processing method involved in the present application.

[0069] See also Figure 7 In an embodiment of the present application, an image processing system 900 is provided, including but not limited to: an acquisition module 910, a registration module 930, a preprocessing module 950, a target detection module 970, an evaluation module 990 and a cropping module 1100.

[0070] An acquisition module 910 is configured to obtain an original image pair II' at different focal plane distances; the original image pair II' includes a first-pass image I and a second-pass image I';

[0071] The registration module 930 is used to register the first path image I with the second path image I' to obtain a registered image pair I R -I';

[0072] The pre-processing module 950 is used to register the image pair I R -I' is preprocessed to obtain the processed image pair I P -I P ';Image pair I P -I P 'Including the first processed image I P and the second processed image I P ';

[0073] The target detection module 970 is used to detect the processed image. P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image;

[0074] Evaluation module 990, for evaluating the first processed image I P The ROI image and the second processed image IP 'The clarity of the ROI images is evaluated respectively, so as to screen out the first coordinate and the second coordinate of the clear ROI image respectively;

[0075] The cropping module 1100 is configured to use the first coordinates of the clear ROI image to respectively crop the first processed image I P and the second processed image I P 'Crop, and then use the second coordinates of the clear ROI image to respectively P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

[0076] It should be noted that, when performing image processing, the image processing system provided in the above embodiment only uses the division of the above functional modules as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the image processing device will be divided into different functional modules to complete all or part of the functions described above.

[0077] In addition, the image processing system and the image processing method provided in the above embodiments belong to the same concept, and the specific manner in which each module performs operations has been described in detail in the method embodiments and will not be repeated here.

[0078] See also Figure 8 In an embodiment of the present application, an image processing device 4000 is provided. The image processing device 4000 may include a desktop computer, a laptop computer, a server, etc.

[0079] exist Figure 8 In the embodiment, the image processing device 4000 includes at least one processor 4001 , at least one communication bus 4002 and at least one memory 4003 .

[0080] The processor 4001 and the memory 4003 are connected, for example, via a communication bus 4002. Optionally, the electronic device 4000 may further include a transceiver 4004, which may be used for data exchange between the image processing device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the number of transceivers 4004 is not limited to one, and the structure of the image processing device 4000 does not constitute a limitation on the embodiments of the present application.

[0081] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0082] The communication bus 4002 may include a path for transmitting information between the above components. The communication bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The communication bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0083] The memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited to these.

[0084] The memory 4003 stores a computer program, and the processor 4001 reads the computer program stored in the memory 4003 through the communication bus 4002 .

[0085] When the computer program is executed by the processor 4001 , the image processing method in the above-mentioned embodiments is implemented.

[0086] In addition, an embodiment of the present application provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the image processing method in the above embodiments is implemented.

[0087] In an embodiment of the present application, a computer program product is provided, which includes a computer program stored in a storage medium. A processor of a computer device reads the computer program from the storage medium, and the processor executes the computer program, so that the computer device performs the image processing method in each of the above embodiments. It is easy for those skilled in the art to understand that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An image processing method based on a dual-channel imaging device, wherein the dual-channel imaging device includes two imaging systems, characterized in that: The method comprises the following steps: Obtaining original image pairs II' at different focal plane distances; the original image pairs II' include a first-channel image I and a second-channel image I'; The first path image I is registered with the second path image I' to obtain a registered image pair I R -I'; For the registered image pair I R -I' is preprocessed to obtain the processed image pair I P -I P '; The image pair I P -I P 'Including the first processed image I P and the second processed image I P '; For the processed image pair I P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image; The first processed image I P The ROI image and the second processed image I P 'The clarity of the ROI images is evaluated respectively, so as to screen out the first coordinate and the second coordinate of the clear ROI image respectively; Utilize the first coordinate of the clear ROI image to respectively obtain the first processed image I P and the second processed image I P ', and use the second coordinates of the clear ROI image to respectively extract the first processed image I P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

2. The image processing method according to claim 1, wherein: The method further comprises: The clear image in the clear-blurred image pair at different defocus distances is cleaned, images that do not meet the set requirements in the clear image are removed, and blurred images corresponding to the removed images in the clear image are removed.

3. The image processing method according to claim 1, wherein: The first path image I is registered with the second path image I' to obtain a registered image pair I R -I', including; Establishing a registration model of the two-path imaging system using the target; The first path image I is registered with the second path image I' using the registration model to obtain a registered image pair I R -I'.

4. The image processing method according to claim 1, wherein: The pre-processing includes background subtraction and white balance.

5. The image processing method according to any one of claims 1 to 4, characterized in that: The method further comprises: Using the defocus distance as the label, a dataset of clear and blurred ROI in situ image pairs is established.

6. The image processing method according to claim 5, wherein: The defocus distance is used as a label, including: When the imaging target is located between the focal plane and the two imaging systems, the defocus distance is equal to the focal plane distance; when the imaging target is located outside the focal plane, the defocus distance is the inverse of the focal plane distance.

7. An image processing system based on a dual-channel imaging device, wherein the dual-channel imaging device comprises two imaging systems, characterized in that: The image processing system comprises: An acquisition module, configured to obtain an original image pair II' at different focal plane distances; the original image pair II' comprises a first-channel image I and a second-channel image I'; The registration module is used to register the first path image I with the second path image I' to obtain a registered image pair I R -I'; A pre-processing module is used to process the registered image pair I R -I' is preprocessed to obtain the processed image pair I P -I P '; The image pair I P -I P 'Including the first processed image I P and the second processed image I P '; The target detection module is used to detect the processed image P -I P 'Perform target detection and extract target ROI image; the target ROI image includes the image from the first processed image I P The ROI image and the second processed image I P 'ROI image; An evaluation module is used to evaluate the first processed image I P The ROI image and the second processed image I P 'The clarity of the ROI images is evaluated respectively, so as to screen out the first coordinate and the second coordinate of the clear ROI image respectively; The cropping module is configured to use the first coordinates of the clear ROI image to respectively crop the first processed image I P and the second processed image I P ', and use the second coordinates of the clear ROI image to respectively extract the first processed image I P and the second processed image I P 'Crop in the middle to obtain clear-blurred image pairs at different defocus distances.

8. An image processing device, characterized in that include: at least one processor, at least one memory, and at least one communication bus, wherein: The memory stores a computer program, and the processor reads the computer program in the memory through the communication bus; When the computer program is executed by the processor, the image processing method according to any one of claims 1 to 6 is implemented.

9. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the image processing method according to any one of claims 1 to 6 is implemented.

Images