Image exposure correction method and program product

Abstract

The invention discloses an image exposure correction method and a program product. The method comprises the following steps: acquiring first image data containing a target object through an image acquisition device, and determining a first image exposure feature according to the first image data; acquiring first object data containing the target object through an image sensor, and determining a second image exposure feature according to the first object data; and determining a third image exposure feature according to the first image exposure feature and the second image exposure feature, and determining target image data after exposure correction according to the first image data and the third image exposure feature. Therefore, through the exposure feature fusion operation of the target image, the color and detail reduction effect of the target image is improved, and the accuracy of face recognition is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to an image exposure correction method and program product. BACKGROUND

[0002] In the field of image processing, exposure control in dynamic scenes has always been one of the key technical challenges. With the rapid development of computer vision, autonomous driving, intelligent monitoring and other applications, higher requirements are put forward for the real-time performance, environmental adaptability and imaging quality of image acquisition. Especially under complex lighting conditions (such as backlight, strong shadow or rapidly changing lighting environment), traditional exposure control methods often have difficulty in balancing global brightness equalization and local detail preservation, resulting in overexposure or underexposure of target objects, which seriously affects the accuracy of subsequent analysis.

[0003] In related technologies, mainstream solutions rely on automatic exposure algorithms of a single image sensor, adjusting exposure parameters by counting image histograms or regional brightness means. However, such methods are easily disturbed by scene content and are difficult to optimize exposure for specific target objects, resulting in loss of key information. For example, in a monitoring scene, if the background sky is too bright, the face region may be underexposed due to the global exposure strategy. Some improved solutions attempt to combine multi-sensor data to assist exposure decision-making, but the feature differences between different modal data can lead to unstable fusion results. For example, although infrared images can compensate for low-light information, their exposure features are difficult to directly align with those of visible light images, which may introduce noise or artifacts. Moreover, existing technologies are mostly designed based on static or slowly changing lighting conditions, and have a lag response to rapidly moving targets or sudden lighting changes, resulting in exposure jumps between consecutive frames and disrupting temporal consistency.

[0004] In summary, there is an urgent need for an exposure control scheme that can efficiently fuse multi-source image features, dynamically adapt to target object requirements and consider real-time performance. SUMMARY

[0005] The present application provides an image exposure correction method and program product to solve the technical problem of poor image display effect in related technologies.

[0006] According to an aspect of the present application, an image exposure correction method is provided, which comprises:

[0007] acquiring first image data containing a target object through an image acquisition device, and determining first image exposure features according to the first image data;

[0008] acquiring first object data containing the target object through an image sensor, and determining second image exposure features according to the first object data;

[0009] A third image exposure characteristic is determined according to the first image exposure characteristic and the second image exposure characteristic, and target image data after exposure correction is determined according to the first image data and the third image exposure characteristic.

[0010] According to another aspect of the present invention, there is provided an image exposure correction device, the device comprising:

[0011] an image data processing module, configured to acquire first image data containing a target object through an image acquisition device, and determine a first image exposure feature based on the first image data;

[0012] an object data processing module, configured to collect first object data containing the target object through an image sensor, and determine a second image exposure feature based on the first object data;

[0013] An image synthesis module is used to determine a third image exposure characteristic based on the first image exposure characteristic and the second image exposure characteristic, and to determine target image data after exposure correction based on the first image data and the third image exposure characteristic.

[0014] According to another aspect of the present invention, an electronic device is provided, comprising:

[0015] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform an image exposure correction method as described in any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement an image exposure correction method according to any embodiment of the present invention when executed.

[0017] According to another aspect of the present invention, an embodiment of the present disclosure further provides a computer program product, including a computer program, which, when executed by a processor, implements an image exposure correction method as described in any one of the embodiments of the present disclosure.

[0018] The technical scheme of the embodiment of the present application uses an image acquisition device to acquire first image data containing a target object, determines a first image exposure feature according to the first image data, so that the image exposure feature data can be directly obtained based on the first image, and the image exposure feature data is efficiently and conveniently obtained. Further, a first object data containing the target object is acquired by using an image sensor, a second image exposure feature is determined according to the first object data, so that the object data related to the target object can be obtained from another dimension by analyzing the original data based on the original data of the target object, and the first object data is subjected to exposure feature extraction, so that the image exposure feature data with richer image content is obtained, and the accuracy and effectiveness of the image exposure feature are enhanced. Finally, a third image exposure feature is determined according to the first image exposure feature and the second image exposure feature, and target image data subjected to exposure correction is determined according to the first image data and the third image exposure feature. The first image data directly acquired is processed in combination with the main image exposure feature data in the target object, so that the target image data obtained finally can contain as many main exposure feature data in the image as possible, so that the accuracy of the target image data generated finally is improved, and the recognition accuracy of the target image is further improved.

[0019] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0021] Figure 1 is a flow chart of an image exposure correction method provided by the first embodiment of the present application;

[0022] Figure 2 is a flow chart of an image exposure correction method provided by the second embodiment of the present application;

[0023] Figure 3A is a schematic diagram of a data processing flow of an image exposure correction method which can be used in the embodiment of the present application and is provided by the third embodiment of the present application;

[0024] Figure 3Bis a schematic diagram of an exposure correction processing flow of an image exposure correction method according to the third embodiment of the present application;

[0025] Figure 4 is a structural schematic diagram of an image exposure correction device according to the third embodiment of the present application;

[0026] Figure 5 is a structural schematic diagram of an electronic device implementing an image exposure correction method according to the third embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the personnel in the technical field better understand the present application scheme, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", "third", "target" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0029] It should be noted that the modification of "one" or "multiple" mentioned in the present disclosure is illustrative but not restrictive, and those skilled in the art should understand that unless otherwise explicitly indicated in the context, it should be understood as "one or more".

[0030] The names of the messages or information exchanged between the multiple devices in the embodiments of the present disclosure are only for illustrative purposes, and are not intended to limit the scope of the messages or information.

[0031] It can be understood that the data involved in the present technical scheme (including but not limited to the data itself, the acquisition or use of the data) should comply with the requirements of the corresponding laws and regulations and relevant provisions.

[0032] Embodiment one

[0033] Figure 1 A flowchart of an image exposure correction method is provided for the first embodiment of the present invention. This embodiment is applicable to scenarios where facial images of target objects are to be recognized. The method can be performed by an image exposure correction device, which can be implemented in the form of hardware and / or software. Optionally, it can be implemented by an electronic device, which can be a mobile terminal, a PC, or a server. Figure 1 As shown, the method may specifically include:

[0034] S110 , capturing first image data containing a target object through an image capturing device, and determining a first image exposure feature based on the first image data.

[0035] The image acquisition device can be understood as a device that converts captured data into an image. For example, the image acquisition device can include at least one of a smartphone camera, a tablet camera, a computer camera, a webcam, an access control or attendance camera, and a 3D scanner. The first image data is image data obtained by capturing an image of a target object using the image acquisition device.

[0036] The target object can be understood as the object that needs to be captured during image acquisition. The target object is usually the object that the user is more concerned about. Its specific content can be set according to actual needs and is not specifically limited here. For example, the target object can be one or more types of objects such as people, animals, plants, buildings, decorations, and furniture. The target object can also refer to the target area for image acquisition, etc. The target object can also be a sub-object of an object, that is, its local area. In some scenarios, the target object can be a preset part, where the preset part includes at least one or more parts such as face, eyes, limbs, and torso.

[0037] The first image exposure characteristic specifically refers to an image exposure characteristic result obtained by encoding the first image data. For example, the image exposure characteristic may include at least one of brightness statistics, highlight and shadow detection results, image detail retention index, noise sensitivity, light source direction, and light source color temperature.

[0038] Specifically, an image acquisition device performs an image acquisition operation on a target object to obtain first image data of the target object. The first image data is then encoded to determine a first image exposure characteristic corresponding to the first image data. By directly performing exposure processing on the image data of the target object, the exposure characteristic of the image data can be quickly acquired, thereby improving image data processing efficiency.

[0039] In another embodiment, determining the first image exposure characteristic based on the first image data includes inputting the first image data into an exposure correction characteristic extraction module to obtain the first image exposure characteristic; wherein the exposure correction characteristic extraction module includes an encoder and a decoder. This technical solution enables global processing of the first image data to achieve global exposure correction for the first image data.

[0040] In another embodiment, determining the first image exposure feature based on the first image data includes: determining the object area image data of the target object based on the first image data, and inputting the object area image data into an exposure correction feature extraction module to obtain the first image exposure feature; wherein the exposure correction feature extraction module includes an encoder and a decoder.

[0041] The object area image data specifically refers to image data of a specific area obtained by performing image detection on a preset part or target area. The exposure correction feature extraction module refers to a functional module for extracting exposure feature data from the object area image data.

[0042] Specifically, by analyzing and detecting the first image data, a target range within the first image data belonging to a preset portion or target area is determined. Then, by processing the image data within the target range, object region image data belonging to the preset portion or target area is determined. Furthermore, by inputting the object region image data into a preset exposure correction feature extraction module, first image exposure features belonging to the preset portion or target area are obtained. By determining the object region image data belonging to the preset portion or target area, this technical solution enables more targeted analysis and processing of image data in specific regions, thereby enabling more refined feature analysis and processing of important image data. This reduces the amount of data that needs to be processed, alleviates data processing pressure, and improves the accuracy and comprehensiveness of image exposure feature acquisition.

[0043] In one embodiment, determining the object area image data of the target object based on the first image data includes: inputting the first image data into a first feature extraction module to obtain the object detection feature of the target object; inputting the object detection feature into an object detection module to obtain the object area image data of the target object.

[0044] Among them, the first feature extraction module refers to a functional module for extracting object detection features located at a preset position or target area in the first image. Specifically, the first feature extraction module can be constructed based on at least one algorithm such as a color histogram algorithm, a convolutional neural network algorithm, a graph neural network algorithm, etc. The object detection feature can be understood as feature data used to characterize the range of a preset position or target area. The object detection module refers to a functional module for performing feature extraction processing based on object detection feature data. Specifically, the object detection module can be constructed based on at least one algorithm such as a threshold segmentation algorithm, an edge segmentation algorithm, a point cloud detection algorithm, etc., and is used to detect the target range of the image.

[0045] In an embodiment of the present invention, the object detection module can be obtained by training a machine learning model based on a sample detection image and its corresponding expected detection image. The expected detection image can be understood as a local image containing the target object segmented from the sample detection image.

[0046] Specifically, the first image data is input into a feature extraction module, and image data within the range of the target object is detected and processed to determine a target range corresponding to a predetermined portion or target area of ​​the target object, i.e., an object detection feature. Furthermore, the acquired object detection feature is input into the object detection module to extract feature data of a specific area based on the object detection feature, thereby obtaining object area image data of the target object. This allows for feature data extraction of specific image areas, thereby improving the accuracy of the image feature data extraction results.

[0047] S120: Collect first object data containing the target object through an image sensor, and determine a second image exposure feature based on the first object data.

[0048] The image sensor specifically refers to a core device that converts optical images into electrical signals. For example, the image sensor may include at least one of a charge-coupled device (CCD) image sensor and a complementary metal oxide semiconductor (CMOS) image sensor.

[0049] First object data specifically refers to the raw data corresponding to the target object acquired by the image sensor. This data is in its original form, unprocessed, converted, or interpreted. In other words, first object data, or RAW data, is the unprocessed raw binary data obtained after the image sensor converts the optical signal into an electrical signal. Each pixel typically contains only one of the RGB components.

[0050] The second image exposure feature can be understood as a plurality of image exposure features corresponding to the target object, obtained by extracting features from the first object data. Exemplarily, the second image exposure feature may include at least one of image exposure features such as brightness statistics, noise sensitivity, light source color temperature, and highlight area information entropy decay rate. It should be noted that the feature types of the first and second image exposure features may be the same or different. Optionally, the feature dimensions of the first and second image exposure features may be the same to facilitate subsequent feature fusion processing.

[0051] Specifically, a preset image acquisition sensor is used to capture first object data of a target object, obtaining raw data of the target object. Image conversion processing is performed on the first object data to obtain image data corresponding to the first object data. Image exposure features from the image data are then extracted to determine second image exposure features corresponding to the first object data. This processing of the first object data can thereby yield richer and more realistic image exposure features, thereby improving subsequent image processing.

[0052] In one embodiment, determining the second image exposure feature based on the first object data includes: performing graphic signal processing on the first object data to obtain second image data, and performing feature extraction on the second image data to obtain the second image exposure feature.

[0053] The image signal processing method may include at least one of image signal processing (ISP), histogram equalization, demosaicing, compressed sensing, and event-based cameras. The second image data may be image data converted from the original data in the first object data, and may include richer and more detailed image data.

[0054] Specifically, the first object data is first preprocessed using bit depth mapping and interpolation algorithms. Then, image processing is performed on the preprocessed first object data using a preset graphics signal processing method to determine second image data corresponding to the first object data. Finally, exposure characteristic data processing is performed on the second image data to obtain second image exposure characteristics. By generating second image data based on the first object data and obtaining the second image exposure characteristic data, this technical solution can effectively obtain more complete and comprehensive image exposure characteristic data based on the original data, thereby improving the accuracy of image generation.

[0055] S130: Determine a third image exposure characteristic according to the first image exposure characteristic and the second image exposure characteristic, and determine target image data after exposure correction according to the first image data and the third image exposure characteristic.

[0056] The third image exposure characteristic refers to an image exposure characteristic derived from a fusion of the first and second image exposure characteristics, and includes multiple image exposure characteristics corresponding to the target object. The target image data can be understood as the first image data after exposure correction has been performed based on the third image exposure data.

[0057] Specifically, the first and second image exposure characteristics are aligned according to their characteristic dimensions, and the first and second image exposure characteristics are then fused to obtain a third image exposure characteristic. Furthermore, exposure correction is performed on the third image exposure characteristic to obtain exposure-corrected third image exposure characteristic data. The exposure-corrected third image exposure characteristic data is then synthesized with the first image data to determine exposure-corrected target image data. This allows exposure correction to be performed only on the area where the target object is located. By deeply combining this image data with the original target object data, the display quality of the target image data is effectively improved, thereby enhancing the accuracy of target image recognition.

[0058] In one embodiment, determining the third image exposure feature based on the first image exposure feature and the second image exposure feature includes: inputting the first image exposure feature and the second image exposure feature into an exposure feature fusion module to obtain the third image exposure feature.

[0059] The exposure feature fusion module is a functional module for fusing the dimensions of the first image exposure feature and the second image exposure feature. Specifically, the exposure feature fusion module can be constructed based on at least one algorithm such as a graph convolutional network algorithm, a multimodal generative adversarial network algorithm, and an autoencoder fusion algorithm.

[0060] In an embodiment of the present invention, the exposure feature fusion module can be obtained by training a machine learning model based on the first sample image feature, the second sample image feature and their corresponding expected fusion image data.

[0061] Specifically, the first and second image exposure characteristics are input into the exposure characteristic fusion module, which then fuses the dimensional data in the first and second image exposure characteristics to obtain a fused third image exposure characteristic. This allows accurate image results to be obtained for a preset location or target area based on the fusion of the exposure characteristics in the directly acquired image data with the exposure characteristics in the image data generated based on the original data, thereby improving the accuracy of the image exposure characteristic data.

[0062] The technical solution of the embodiments of the present invention utilizes an image acquisition device to acquire first image data containing a target object, and determines a first image exposure characteristic based on the first image data. This allows image exposure characteristic data to be directly acquired based on the first image, thereby efficiently and conveniently obtaining the image exposure characteristic data. Furthermore, an image sensor is used to acquire first object data containing the target object, and a second image exposure characteristic is determined based on the first object data. This allows object data related to the target object to be acquired from another dimension by analyzing the original data of the target object, thereby extracting exposure characteristics from the first object data. This results in image exposure characteristic data with richer image content, enhancing the accuracy and effectiveness of the image exposure characteristics. Finally, a third image exposure characteristic is determined based on the first and second image exposure characteristics, and target image data after exposure correction is determined based on the first and third image exposure characteristics. This allows the directly acquired first image data to be processed in combination with the primary image exposure characteristic data of the target object, ensuring that the resulting target image data contains as much of the primary exposure characteristic data as possible, thereby improving the accuracy of the resulting target image data and, consequently, the recognition accuracy of the target image.

[0063] Example 2

[0064] Figure 2 This is a flowchart of an image exposure correction method provided in the second embodiment of the present invention. The solution in this embodiment is based on the above embodiment. It refines the technical solution of determining the target image data after exposure correction based on the first image data and the third image exposure characteristics. For specific implementation methods, please refer to the description of this embodiment. Among them, the technical features that are the same or similar to the above embodiments are not repeated here. Figure 2 As shown, the method may specifically include:

[0065] S210: Capture first image data containing a target object through an image acquisition device, and determine a first image exposure feature based on the first image data.

[0066] S220: Collect first object data containing the target object through an image sensor, and determine a second image exposure feature based on the first object data.

[0067] S230: Determine a third image exposure feature based on the first image exposure feature and the second image exposure feature, and map the third image exposure feature into an image space to obtain regional exposure-corrected image data of the object region image data.

[0068] The image space is used to implement exposure correction for a preset part or target area. The regional exposure-corrected image data specifically refers to image data that has undergone exposure correction.

[0069] Specifically, the first and second image exposure characteristics are aligned according to their characteristic dimensions, and then fused to obtain a third image exposure characteristic. The third image exposure characteristic is then mapped into image space, and exposure correction is performed on the third image exposure characteristic within the image space to obtain regional exposure-corrected image data for the object area. This facilitates sensor acquisition of image data information, highlights image details, and effectively improves the overall quality and usability of the image data.

[0070] In one embodiment, mapping the third image exposure feature into the image space to obtain regional exposure correction image data of the object area image data includes: inputting the third image exposure feature into an exposure image generation module to map the third image exposure feature into the image space to obtain regional exposure correction image data of the object area image data.

[0071] The exposure image generation module is a functional module for performing exposure correction on exposure feature data. Specifically, the exposure image generation module can be constructed based on at least one of a feature-aligned multi-image exposure generation algorithm, a reference image exposure generation algorithm, and a multimodal fusion exposure generation algorithm.

[0072] In an embodiment of the present invention, the exposure image generation module can be obtained by training a machine learning model based on the exposure characteristics of the sample image and its corresponding expected correction image data.

[0073] Specifically, the third image exposure feature is input into the exposure image generation module, thereby mapping the third image exposure feature into image space. Exposure correction is then performed on a predetermined portion or target area within the third image exposure feature within the image space, thereby generating regional exposure-corrected image data for the target area image data. This facilitates exposure correction processing of the image feature data, resulting in accurately exposed image data.

[0074] S230: Generate exposure-corrected target image data according to the first image data and the regional exposure-corrected image data.

[0075] The target image data refers to image data obtained by splicing the regional exposure-corrected image data and the first image data.

[0076] Specifically, by splicing the regional exposure-corrected image data with the feature data of a preset location or target area in the first image data, the first image data, updated and adjusted based on the regional exposure-corrected image data, is generated, i.e., the target image data. This allows the target image data to contain more comprehensive, detailed, accurate, and important image data, effectively improving the quality of the target image and ensuring the accuracy of target image recognition.

[0077] In one embodiment, generating exposure-corrected target image data based on the first image data and the regional exposure-corrected image data includes: inputting the first image data and the regional exposure-corrected image data into an image synthesis module to obtain exposure-corrected target image data.

[0078] The image synthesis module refers to a functional module for performing image synthesis. Optionally, the image synthesis module can be constructed based on at least one of an image segmentation synthesis algorithm, a texture synthesis algorithm, a diffusion model synthesis algorithm, and a flow model synthesis algorithm. The preset method in the image stitching synthesis processing can also include at least one of a data fusion and stitching method such as linear fusion, multi-band fusion, and Poisson fusion.

[0079] In an embodiment of the present invention, the image synthesis module may be obtained by training a machine learning model based on the first sample image, the second sample image, and their corresponding expected synthesized image. The second sample image may be an image that includes partial image content from the first sample image. Specifically, the second sample image may be a partial image of the first sample image, or an image obtained by processing a partial image of the first sample image.

[0080] Specifically, the first image data and the regional exposure-corrected image data are input into the image synthesis module, which then performs splicing and synthesis processing on the first image data and the regional exposure-corrected image data according to a preset method to obtain the target image data after exposure correction. This effectively improves the quality of the directly captured image, allowing the target image data to more comprehensively and accurately display the image content, thereby improving image recognition accuracy.

[0081] The technical solution of this embodiment of the present invention first maps the third image exposure feature into image space to obtain regional exposure-corrected image data for the object region image data, thereby obtaining high-quality exposure-corrected image data and improving the image quality of the regional exposure-corrected image data. Furthermore, exposure-corrected target image data is generated based on the first image data and the regional exposure-corrected image data. This allows the first image data to be deeply integrated with the feature data in the regional exposure-corrected image data, ensuring that the target image data contains comprehensive, detailed, and accurate image data, thereby improving the display quality of the target image.

[0082] Example 3

[0083] The third embodiment of the present invention provides a flowchart of an image exposure correction method. In order to better introduce the technical solution provided by the embodiment of the present invention, this embodiment takes the image exposure correction processing of a facial image as an example. The flowchart of this embodiment is as follows: Figure 3A and Figure 3B As shown, the technical features that are the same or similar to those in the aforementioned embodiments are not repeated here.

[0084] like Figure 3A As shown, the implementation framework of the image exposure correction method according to the embodiment of the present invention may specifically include:

[0085] 1) Image acquisition unit (device):

[0086] The image acquisition unit consists of an image sensor and an image preprocessing module. The image sensor is used to convert optical signals into electrical signals to obtain RAW data. The image preprocessing module processes the RAW data into an RGB image using the image acquisition unit's built-in ISP algorithm. In this example, the image sensor can be integrated into the image acquisition device.

[0087] 2) Target detection unit (i.e. object detection module):

[0088] The target detection unit consists of a facial feature extraction module and a facial detection module. The former extracts the features required for facial detection through a neural network, and the latter performs facial region detection and outputs a facial region image. The specific flow chart is as follows Figure 2 shown.

[0089] 3) Exposure correction unit:

[0090] The unit is composed of an exposure correction feature extraction module 3.1, an exposure correction feature fusion module 3.2, and an exposure output module 3.3. The exposure correction feature extraction module 3.1 is based on an encoder-decoder structure and realizes extraction of exposure features in the RGB image directly collected by the image device module; the ISP image exposure correction extraction module 4.3 is used for exposure correction feature extraction on the ISP-processed RGB image data; the exposure correction feature fusion module 3.2 uses a multi-layer convolutional network to combine the exposure features in the RGB image and the exposure correction features of the ISP-processed image data to realize feature fusion; and the exposure output module 3.3 maps the fused features from the feature space to the image space to realize exposure correction of the face region image.

[0091] 4) ISP unit:

[0092] The unit is composed of a RAW data preprocessing module 4.1, an ISP image signal processing module 4.2, and an ISP image exposure correction extraction module 4.3 for exposure correction feature extraction on the ISP-processed RGB image data. The RAW data preprocessing module 4.1 reshapes, maps the bit depth, and performs interpolation algorithm processing on the RAW data; the ISP image signal processing module 4.2 performs AWB white balance, CCM color correction, and Gamma curve adjustment processing on the RAW data; and the ISP image exposure correction extraction module 4.3 is used for exposure correction feature extraction on the ISP-processed RGB image data.

[0093] 5) Image synthesis unit:

[0094] The unit is composed of an image synthesis module 5.1. The image synthesis module 5.1 is used for splicing the exposure-corrected face region image and the RGB image output by the image acquisition unit to obtain the final output image.

[0095] As shown in Figure 3B , the implementation process of the image exposure correction method of the embodiment of the present application can specifically include:

[0096] Step one: acquire the RGB image containing the face through the image acquisition module (such as a camera). Then the RGB image can be preprocessed, for example, filtering processing and / or image enhancement processing, etc. The preprocessed RGB image is detected to detect the face region and output the face region image. The face region image is extracted by the pre-set face feature extraction module to extract the image features required for face exposure correction. Then, the face region image is processed by the exposure correction feature extraction module to extract the exposure features corresponding to the face region image, i.e., the first image exposure features.

[0097] Step 2: Use the image sensor to capture the light signal of the face and convert it into an electrical signal to obtain RAW data. RAW data is the unprocessed original data directly captured by the image sensor, which retains complete information such as light intensity and color for subsequent ISP optimization. The raw data (RAW data) collected by the image sensor is preprocessed by the ISP image signal processing unit. For example, the RAW data can be processed into a facial RGB image by one or more processes such as denoising, white balancing, color correction, and edge enhancement. Then, the facial RGB image obtained by processing the RAW data is converted into a second image exposure feature of the same dimension as the first image exposure feature through an exposure correction feature extraction module (such as a downsampling module).

[0098] Step 3: The exposure correction feature fusion module fuses the first and second image exposure features to generate fused image data, namely, the third image exposure feature. The exposure output module then maps the third image exposure feature space to the image space to perform exposure correction on the facial region image, thereby generating exposure-corrected image data, namely, the regional exposure-corrected image data.

[0099] Step 4: Splice the regional exposure-corrected image data with the RGB image to obtain exposure-corrected image data, and output the final image result, i.e., the target image data, to obtain a facial image with a better exposure correction effect.

[0100] The technical solution in the embodiments of the present invention designs a novel exposure correction scheme for facial images by combining target detection, deep learning, and the ISP algorithm. Exposure correction is performed after locating the facial area through target detection, avoiding unnecessary computing costs by performing exposure correction on the entire image. This fully utilizes the rich information in the RAW data, unlike using only RGB images as input and losing the collected original information. Furthermore, by combining the feature extraction capabilities of the deep learning model with the image optimization advantages of the ISP algorithm, the fusion effect of the exposure features is enhanced, further improving the quality and details of the facial image. Furthermore, the exposure information of different areas in the facial image is adaptively adjusted by the exposure correction unit, achieving color and detail restoration of the face under different lighting scenarios.

[0101] Example 4

[0102] Figure 4 This is a schematic diagram of the structure of an image exposure correction device provided by the fourth embodiment of the present invention. The device can be implemented by software and / or hardware and can be configured in an electronic device. Figure 4 As shown, the device includes: an image data processing module 401, an object data processing module 402 and an image synthesis module 403.

[0103] Among them, the image data processing module 401 is used to collect first image data containing the target object through an image acquisition device, and determine a first image exposure characteristic based on the first image data; the object data processing module 402 is used to collect first object data containing the target object through an image sensor, and determine a second image exposure characteristic based on the first object data; the image synthesis module 403 is used to determine a third image exposure characteristic based on the first image exposure characteristic and the second image exposure characteristic, and determine the target image data after exposure correction based on the first image data and the third image exposure characteristic.

[0104] According to the technical solution of the embodiment of the present invention, the image data processing module 401 uses an image acquisition device to acquire first image data containing a target object and determines a first image exposure characteristic based on the first image data. This allows the image exposure characteristic data to be directly acquired based on the first image, thereby efficiently and conveniently obtaining the image exposure characteristic data. Furthermore, the object data processing module 402 uses an image sensor to acquire first object data containing the target object and determines a second image exposure characteristic based on the first object data. This allows the object data related to the target object to be acquired from another dimension by analyzing the original data of the target object, thereby extracting exposure characteristics from the first object data. This results in image exposure characteristic data with richer image content, thereby enhancing the accuracy and effectiveness of the image exposure characteristics. Finally, the image synthesis module 403 determines a third image exposure characteristic based on the first and second image exposure characteristics, and determines the target image data after exposure correction based on the first and third image exposure characteristics. Therefore, the first image data directly collected is processed in combination with the main image exposure feature data in the target object, so that the final target image data can contain as much main exposure feature data in the image as possible, thereby improving the accuracy of the final generated target image data and further improving the recognition accuracy of the target image.

[0105] Based on the above optional technical solutions, the image data processing module 401 may optionally further include a first image exposure feature determination unit. The first image exposure feature determination unit is configured to determine object area image data of the target object based on the first image data, and input the object area image data into an exposure correction feature extraction module to obtain a first image exposure feature. The exposure correction feature extraction module includes an encoder and a decoder.

[0106] Based on the above optional technical solutions, optionally, the first image exposure feature determination unit further includes: an object area image data determination subunit and a first image exposure feature determination subunit. The object area image data determination subunit is configured to input the first image data into a first feature extraction module to obtain object detection features of the target object. The first image exposure feature determination subunit is configured to input the object detection features into an object detection module to obtain object area image data of the target object.

[0107] Based on the above optional technical solutions, the object data processing module 402 may optionally further include a second image exposure feature determination unit. The second image exposure feature determination unit is configured to perform graphic signal processing on the first object data to obtain second image data, and perform feature extraction on the second image data to obtain a second image exposure feature.

[0108] Based on the above optional technical solutions, the image synthesis module 403 may optionally further include a third image exposure feature determination unit, wherein the third image exposure feature determination unit is configured to input the first image exposure feature and the second image exposure feature into an exposure feature fusion module to obtain a third image exposure feature.

[0109] Based on the above optional technical solutions, the image synthesis module 403 optionally further includes: a regional exposure correction image data determination unit and a target image data determination unit. The regional exposure correction image data determination unit is configured to map the third image exposure feature into image space to obtain regional exposure correction image data for the object region image data. The target image data determination unit is configured to generate target image data after exposure correction based on the first image data and the regional exposure correction image data.

[0110] Based on the above optional technical solutions, optionally, the regional exposure correction image data determination unit further includes a regional exposure correction image data processing subunit. The regional exposure correction image data processing subunit is configured to input the third image exposure feature into an exposure image generation module to map the third image exposure feature into an image space, thereby obtaining regional exposure correction image data for the object region image data.

[0111] Based on the above-mentioned optional technical solutions, optionally, the target image data determination unit also includes: a target image data determination subunit, which is used to input the first image data and the regional exposure correction image data into the image synthesis module to obtain the target image data after exposure correction.

[0112] Based on the above optional technical solutions, optionally, the target object includes a preset part, and the preset part includes at least a face.

[0113] The image exposure correction device provided in the embodiments of the present invention can implement any of the image exposure correction methods provided in any of the embodiments of the present invention, and includes the corresponding functional modules and beneficial effects of the implementation method. For technical details not fully described in this embodiment, please refer to any of the image exposure correction methods described in the embodiments of the present invention.

[0114] Example 5

[0115] Figure 5 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0116] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0117] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0118] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the image exposure correction method.

[0119] In some embodiments, the image exposure correction method may be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the image exposure correction method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the image exposure correction method in any other suitable manner (e.g., via firmware).

[0120] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0121] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0122] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0123] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0124] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0125] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0126] In particular, the processes described above with reference to the flow charts can be implemented as computer software programs in accordance with embodiments of the application. For example, embodiments of the application include a computer program product comprising a computer program carried on a non-transitory computer readable medium, the computer program comprising program code for executing the methods illustrated by the flow charts. In such embodiments, the computer program can be downloaded and installed from a network via the communication unit 19, or installed from the storage unit 18, or installed from the ROM 12. When the computer program is executed by the processor 11, the above-described functions defined in the methods of the embodiments of the application are performed.

[0127] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present application. For example, the steps recited in the present application can be performed in parallel, in series, or in a different order, without departing from the desired results of the technical solutions of the present application, and this is not limited herein.

[0128] The above detailed description does not constitute a limitation on the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for image exposure correction, characterized in that: include: collecting first image data containing the target object by an image acquisition device, and determining a first image exposure feature based on the first image data; collecting first object data containing the target object through an image sensor, and determining a second image exposure feature based on the first object data; A third image exposure characteristic is determined according to the first image exposure characteristic and the second image exposure characteristic, and target image data after exposure correction is determined according to the first image data and the third image exposure characteristic.

2. The image exposure correction method according to claim 1, wherein: The determining of the first image exposure feature according to the first image data includes: The object area image data of the target object is determined based on the first image data, and the object area image data is input into an exposure correction feature extraction module to obtain a first image exposure feature; wherein the exposure correction feature extraction module includes an encoder and a decoder.

3. The image exposure correction method according to claim 2, wherein: The determining the object area image data of the target object according to the first image data includes: inputting the first image data into a first feature extraction module to obtain object detection features of the target object; The object detection features are input into an object detection module to obtain object region image data of the target object.

4. The image exposure correction method according to claim 1, wherein: The determining of the second image exposure feature according to the first object data comprises: Graphic signal processing is performed on the first object data to obtain second image data, and feature extraction is performed on the second image data to obtain second image exposure features.

5. The image exposure correction method according to claim 1, wherein: The determining of the third image exposure feature according to the first image exposure feature and the second image exposure feature includes: The first image exposure feature and the second image exposure feature are input into an exposure feature fusion module to obtain a third image exposure feature.

6. The image exposure correction method according to claim 2, wherein: The determining, based on the first image data and the third image exposure characteristic, target image data after exposure correction includes: Mapping the third image exposure feature into an image space to obtain regional exposure-corrected image data of the object region image data; Exposure-corrected target image data is generated based on the first image data and the regional exposure-corrected image data.

7. The image exposure correction method according to claim 6, wherein: Mapping the third image exposure feature into an image space to obtain regional exposure-corrected image data of the object region image data includes: The third image exposure feature is input into an exposure image generation module to map the third image exposure feature into an image space to obtain regional exposure correction image data of the object region image data.

8. The image exposure correction method according to claim 6, wherein: Generating exposure-corrected target image data according to the first image data and the regional exposure-corrected image data includes: The first image data and the regional exposure-corrected image data are input into an image synthesis module to obtain target image data after exposure correction.

9. The image exposure correction method according to claim 1, wherein: The target object includes a preset part, and the preset part includes at least a face.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the computer program implements the image exposure correction method according to any one of claims 1 to 9.

Images