Method and system for enhancing image quality

Abstract

A method for enhancing image quality can be provided. The method can include receiving a plurality of input frames and metadata, and determining one or more feature scores for a received input frame from the plurality of input frames. The method can also include determining a parameter score for the received input frame based on an analysis of the one or more feature scores for the received input frame and the metadata. The method can include identifying one or more artifacts in the received input frame for correction based on the parameter score, and determining a correction strength required for the at least one identified artifact in the received input frame based on the parameter score, and then applying the determined correction strength to the received input frame. The method can also include performing multi-frame blending on the plurality of received input frames to which the determined correction strength is applied.

Description

Technical Field

[0001] This invention relates generally to image processing, and more specifically to a system and method for enhancing image quality. Background Technology

[0002] Multi-frame blending is a well-known technique for improving the quality of images captured in low-light and extremely low-light conditions. However, the blended output is often affected by multiple artifacts due to noise in the input frames. Since multi-frame blending uses burst capture frames as input, all well-known methods apply preprocessing methods (such as denoising, edge enhancement, etc.) to all input frames before passing them to multi-frame blending. These solutions apply the same settings to all input frames, regardless of their noise / feature characteristics. Some frames may have more noise, some may have better detail and less noise, some may have higher brightness, etc.

[0003] Figure 1 A flowchart 100 depicting a conventional method for multi-frame blending of multiple input frames is shown. In this conventional method, multiple bursts of raw noisy image frames are received by at least one image sensor 101 for multi-frame blending. It then performs frame preprocessing 125 on the input frames using appropriate frame preprocessing techniques to denoise or sharpen each received input frame. Subsequently, the multi-frame blending 105 process may include reference frame selection 110, image registration 115, and blending 120. However, the blended output 130 exhibits artifacts such as noise, blurred edges, and color artifacts. Therefore, a method is needed to reduce the workload / processing required to remove anomalies generated through multi-frame blending. Summary of the Invention

[0004] Solution to the problem

[0005] This overview is provided to introduce some concepts in a simplified form, which will be further described in the detailed description of the invention. This overview is not intended to identify key or fundamental inventive concepts of the invention, nor is it intended to define the scope of the invention.

[0006] According to embodiments of this disclosure, a method for enhancing image quality can be provided. The method may include receiving a plurality of input frames and metadata from an image sensor. The method may include determining one or more feature scores of the received input frames from the plurality of input frames. The method may include determining a parameter score of the received input frames based on analysis of the determined one or more feature scores and the metadata. The method may include identifying one or more artifacts in the received input frames for correction based on the parameter scores. The method may include determining a required correction intensity for at least one identified artifact in the received input frames based on the parameter scores. The method may include applying the determined correction intensity to the received input frames. The method may include performing multi-frame blending on the plurality of received input frames to which the determined correction intensity has been applied.

[0007] The method may include: using context data to extract at least one feature of the received input frame, wherein the context data is determined based on one or more image capture conditions, including ISO, exposure time, and lighting conditions for each frame; and calculating a score for the extracted at least one feature based on the level of the extracted at least one feature of the received input frame.

[0008] The method may further include generating a vector score of at least one of the one or more feature scores of the received input frame based on a weighted average of the one or more feature scores of the received input frame; mixing the generated vector scores of the one or more feature scores of the received input frame; scaling the mixed vector scores; and associating one or more feature vector scores with each of one or more generated feature vectors based on the scaled scores of the mixed vector scores.

[0009] The method may further include: estimating the quality of at least one feature among one or more features of the received input frame based on the parameter score, wherein the quality is estimated based on at least one of peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), or multi-scale structural similarity (MS-SSIM) rating of perceived quality; prioritizing the one or more features based on the estimated quality of the received input frame; determining the correction strength of at least one artifact in the received input frame based on the estimated quality and the priority ranking; and applying the determined correction strength to the received input frame based on the parameter score.

[0010] The method may further include adjusting the parameter fractions of the received input frames to produce optimal results under low-light conditions.

[0011] The method may further include updating the parameter score using changes in image capture conditions; and controlling the parameter score based on the metadata.

[0012] According to an embodiment, the metadata may include at least one of the image sensor's ISO value, BV value, and exposure value. Attached Figure Description

[0013] The above and other aspects, features and advantages of certain embodiments of the present disclosure will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:

[0014] Figure 1 A flowchart illustrating a multi-frame blending process based on existing technology is shown.

[0015] Figure 2 A block diagram of a system for enhancing image quality according to an embodiment of this subject is shown;

[0016] Figure 3 A flowchart depicting a method for enhancing image quality according to an embodiment of this subject is shown;

[0017] Figure 4 A flowchart illustrating an exemplary embodiment depicting a multi-frame blended scene according to embodiments of this subject is shown;

[0018] Figure 5a and Figure 5b A flowchart illustrating an exemplary embodiment of weak light denoising using the proposed solution is shown, according to embodiments of this subject matter;

[0019] Figure 6a and Figure 6b A flowchart illustrating an exemplary embodiment of sharpness-based partial frame denoising using the proposed solution is shown in the embodiments of this subject matter.

[0020] Figure 7a and Figure 7b A diagram illustrating another exemplary embodiment of sharpness-based partial frame denoising using the proposed solution, according to embodiments of this subject matter, is shown.

[0021] Figure 8a and Figure 8b A diagram illustrating yet another exemplary embodiment of sharpness-based partial frame denoising using the proposed solution, according to embodiments of this subject matter; and

[0022] Figure 9a and Figure 9bA diagram illustrating yet another exemplary embodiment of sharpness-based partial frame denoising using the proposed solution, according to embodiments of this subject matter, is shown. Detailed Implementation

[0023] First, it should be understood that although illustrative implementations of embodiments of this disclosure are shown below, the invention can be implemented using any number of techniques (whether currently known or existing). This disclosure should not in any way be limited to the illustrative implementations, drawings, and techniques shown below, including the exemplary designs and implementations shown and described herein, but modifications can be made within the full scope of the appended claims and their equivalents.

[0024] As used herein, the term "some" is defined as "none, one, more than one, or all". Therefore, the terms "none", "one", "more than one", "more than one but not all", or "all" all fall under the definition of "some". The term "some embodiments" can refer to no embodiment, one embodiment, several embodiments, or all embodiments. Therefore, the term "some embodiments" is defined as meaning "no embodiment, one embodiment, more than one embodiment, or all embodiments".

[0025] The terminology and structure used herein are intended to describe, teach, and illustrate certain embodiments and their specific features and elements, and do not limit, constrain, or reduce the spirit and scope of the claims or their equivalents.

[0026] More specifically, unless otherwise stated, any terms used herein, such as, but not limited to, “including,” “contains,” “having,” “comprising,” and their grammatical variations, do not specify exact limitations or constraints, and certainly do not exclude the possibility of adding one or more features or elements, and furthermore, unless otherwise stated in restrictive language such as “must include” or “requires inclusion,” it should not be considered as excluding the possibility of removing one or more listed features and elements.

[0027] Whether or not a feature or element is restricted to use only once, it may still be referred to as "one or more features," "one or more elements," "at least one feature," or "at least one element." Furthermore, unless otherwise specified by restrictive language (such as "requires one or more..." or "requires one or more elements"), the use of the terms "one or more" or "at least one" feature or element does not preclude the absence of that feature or element.

[0028] In the description, the terms “A or B”, “at least one of A or B”, “at least one of A and B”, or “one or more of A and / or B” can include all possible combinations of the items listed together. For example, the terms “A or B” or “at least one of A and / or B” can mean (1) only A, (2) only B, or (3) both A and B.

[0029] As used herein, the expressions “1”, “2”, “first” or “second” may modify various elements regardless of their order and / or importance, and distinguish one element from another without limiting the corresponding element.

[0030] Unless otherwise defined, all terms used herein, in particular any technical and / or scientific terms, may be considered to have the same meaning as commonly understood by one of ordinary skill in the art.

[0031] Embodiments of the present invention will now be described in detail with reference to the accompanying drawings. Embodiments of this disclosure relate to systems for enhancing image quality. Figure 2 A block diagram 200 of a system 202 for enhancing image quality according to an embodiment of this subject is shown. In the embodiment, system 202 may be incorporated into a user equipment (UE). Examples of UEs may include, but are not limited to, televisions, laptops, tags, smartphones, and personal computers (PCs). The system improves the perceptual quality of the resulting mixed image from multi-frame processing by analyzing input frames and transforming the input frames using parameter scores of input features / frame characteristics. Details of the above-described aspects performed by system 202 will be explained below.

[0032] The system may include a processor 204, a memory 206, a module 208, an image sensor 210, an input frame analysis module 212, a transformation module 214, and a multi-frame mixing module 216. In embodiments, the processor 204, memory 206, module 208, image sensor 210, input frame analysis module 212, transformation module 214, and multi-frame mixing module 216 may be communicatively coupled to each other. At least one of the multiple modules 208 may be implemented using an AI model. AI-related functions may be executed using non-volatile memory or volatile memory and / or a processor.

[0033] Processor 204 may include one or more processors. In this case, the one or more processors may be general-purpose processors (such as central processing unit (CPU), application processor (AP), etc.), pure graphics processing units (such as graphics processing unit (GPU)), vision processing unit (VPU) and / or AI-specific processors (such as neural processing unit (NPU)).

[0034] Multiple processors control the processing of input data according to predefined operating rules or artificial intelligence (AI) models stored in non-volatile or volatile memory. The predefined operating rules or AI models are provided through training or learning. Here, "provided through learning" means applying learning techniques to multiple learning data to form predefined operating rules or AI models with desired characteristics. Learning can be performed on the device itself that performs the AI ​​according to the embodiment, and / or can be implemented through a separate server / system. The AI ​​model can consist of multiple neural network layers. Each layer has multiple weight values, and layer operations are performed through computation of the previous layer and operations on the multiple weights. Examples of neural networks include, but are not limited to, convolutional neural networks (CNNs), deep neural networks (DNNs), recurrent neural networks (RNNs), restricted Boltzmann machines (RBMs), deep belief networks (DBNs), bidirectional recurrent deep neural networks (BRDNNs), generative adversarial networks (GANs), and deep Q-networks.

[0035] Learning techniques are methods for training a predetermined target device (e.g., a robot) using multiple learning data sets to enable, allow, or control the target device to make determinations or predictions. Examples of learning techniques include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0036] According to this topic, in methods for electronic devices, there is a method for enhancing image quality. An artificial intelligence model can be obtained through training. Here, "obtained through training" means obtaining a predefined operating rule or artificial intelligence model configured to perform desired features (or purposes) by training a basic artificial intelligence model with multiple training data using training techniques. The artificial intelligence model may include multiple neural network layers. Each of the multiple neural network layers may include multiple weight values, and neural network computation is performed through calculations between the computation results of the previous layer and the multiple weight values.

[0037] Visual understanding is a technology used to identify and process things like human vision, and can include, for example, object recognition, object tracking, image retrieval, human recognition, scene recognition, 3D reconstruction / localization, or image enhancement.

[0038] As will be understood, system 202 can be understood as one or more of hardware, software, logic-based programs, configurable hardware, etc. In the example, processor 204 can be a single processing unit or multiple units, all of which can include multiple computing units. Processor 204 can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, processor cores, multi-core processors, multiprocessors, state machines, logic circuits, application-specific integrated circuits, field-programmable gate arrays, and / or any means of manipulating signals based on operating instructions. Among other capabilities, processor 204 can be configured to acquire and / or execute computer-readable instructions and / or data stored in memory 206.

[0039] In this example, memory 206 may include any non-transitory computer-readable medium known in the art, such as volatile memory (e.g., static random access memory (SRAM) and / or dynamic random access memory (DRAM)) and / or non-volatile memory (e.g., read-only memory (ROM), erasable programmable ROM (EPROM), flash memory, hard disk, optical disk, and / or magnetic tape). Memory 206 may include data. Among other things, the data serves as a repository for storing data processed, received, and generated by one or more of the processor 204, memory 206, module 208, image sensor 210, input frame analysis module 212, transformation module 214, and multi-frame mixing module 216.

[0040] Among other things, module 208 may include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement data types. Module 208 may also be implemented as a signal processor, state machine, logic circuit, and / or any other means or component that manipulates signals based on operation instructions. Furthermore, module 208 may be implemented in hardware as instructions executed by at least one processing unit (e.g., processor 204) or a combination thereof. The processing unit may be a general-purpose processor that executes instructions to cause a general-purpose processor to perform operations, or the processing unit may be dedicated to performing the desired function. In another aspect of this disclosure, module 208 may be machine-readable instructions (software) that, when executed by the processor / processing unit, can perform any of the described functions. In some example embodiments, module 208 may be machine-readable instructions (software) that, when executed by the processor 204 / processing unit, performs any of the described functions.

[0041] In one embodiment, processor 204 may be configured to receive multiple input frames with metadata from image sensor 210. Image sensor 210 may be configured to capture images with bursts of raw noise. The metadata includes at least one of the image sensor 210's ISO value, BV value, and exposure value.

[0042] The input frame analysis module 212 may include a feature scoring module 218 and a parameter scoring module 220. The input frame analysis module 212 may be configured to determine one or more feature scores of a plurality of received input frames. The input frame analysis module 212 may be configured to extract one or more features of the received input frames using context data. Furthermore, the context data may be determined based on one or more image capture conditions, including ISO per frame, exposure time, and lighting conditions. The input frame analysis module 212 may be configured to calculate a score for at least one extracted feature based on the level of the extracted feature of the received input frame. For example, the input frame analysis module 212 may be configured to calculate sharpness scores, noise scores, luminance scores, and other similar feature scores. The input frame analysis module 212 may be configured to determine parameter scores of the received input frames based on analysis of the determined one or more feature scores and the metadata of the received input frames.

[0043] In an embodiment, the input frame analysis module 212 for determining parameter scores can be configured to generate a vector score of at least one feature score of the received input frame based on a weighted average of one or more feature scores of the received input frame. The input frame analysis module 212 can be configured to blend the generated vector scores of one or more feature scores of the received input frame. The input frame analysis module 212 can be configured to scale the blended generated vector scores. The input frame analysis module 212 can be configured to associate one or more generated feature vectors with each of the one or more generated feature vectors based on the scaled scores of the blended generated vectors.

[0044] In an embodiment, the transformation module 214 can be configured to identify one or more artifacts in the received input frame for correction based on parameter scores. The transformation module 214 can be configured to estimate the quality of at least one feature of the received input frame based on the parameter scores. The quality can be estimated based on at least one of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), or Multi-Scale Structural Similarity (MS-SSIM) rating of perceived quality. The transformation module 214 can be configured to prioritize one or more features based on the estimated quality of the received input frame.

[0045] In an embodiment, the transformation module 214 can be configured to determine the correction intensity of at least one artifact in the received input frame based on estimated quality and priority characteristics. The transformation module 214 can be configured to apply the determined artifact correction intensity to the corresponding input frame based on the parameter score. The transformation module 214 can be configured to determine the required correction intensity for at least one artifact identified in at least one received input frame based on the parameter score. The transformation module 214 can be configured to apply the determined intensity to artifact correction in each input frame. The transformation module 214 can be configured to adjust the parameter score of the received input frame to produce optimal results under low-light conditions. The transformation module 214 can be configured to update the parameter score using changes in image capture conditions and to control the parameter score based on metadata.

[0046] Subsequently, the multi-frame mixing module 216 may include a reference frame selection module 222, an image registration module 224, and a mixing module 226. The multi-frame mixing module 216 can be configured to perform multi-frame mixing on the artifact-corrected input frames.

[0047] Figure 3 A flowchart 300 depicting a method for enhancing image quality according to an embodiment of this subject is shown. Method 300 may include receiving, at operation 301, a plurality of input frames having metadata from an image sensor 210. The metadata may include at least one of the camera sensor's ISO value, BV value, and exposure value.

[0048] In an embodiment, method 300 may include determining one or more feature scores of a plurality of received input frames at operation 303 via input frame analysis module 212. Determining one or more feature scores may include extracting one or more features of the received input frames using context data, and calculating a score for the extracted at least one feature based on the level of the extracted at least one feature of the received input frames. Furthermore, the context data may be determined based on one or more image capture conditions including ISO per frame, exposure time, and lighting conditions.

[0049] In an embodiment, method 300 may include, at operation 305, an input frame analysis module 212 determining a parameter score of the received input frame based on analysis of one or more determined feature scores and metadata of the received input frame. Determining the parameter score may include generating a vector score of at least one of the feature scores of the received input frame based on a weighted average of one or more feature scores of the received input frame, mixing the generated vector scores of the one or more feature scores of the received input frame, scaling the mixed generated vector score, and associating one or more generated feature vectors with each of the one or more generated feature vectors based on the scaled score of the mixed generated vector.

[0050] In an embodiment, method 300 may include, at operation 307, the input frame analysis module 212 identifying one or more artifacts in the received input frame for correction based on parameter scores. The method may include estimating the quality of at least one feature of the received input frame based on the parameter scores. The quality may be estimated based on at least one of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), or Multi-Scale Structural Similarity (MS-SSIM) rating of perceived quality. The method may include prioritizing one or more features based on the estimated quality of the received input frame. The method may include determining the correction strength of at least one artifact in the received input frame by the input frame analysis module 212 based on the estimated quality and priority features. The method may include applying the determined artifact correction strength in the corresponding input frame by the input frame analysis module 212 based on the parameter scores.

[0051] In an embodiment, method 300 may include, at operation 309, a transformation module 214 determining, based on a parameter score, the required correction intensity for at least one artifact identified in at least one received input frame.

[0052] In an embodiment, method 300 may include, at operation 311, the transformation module 214 applying a determined artifact correction intensity in each input frame. The method may include, by the transformation module 214, adjusting the parameter fractions of the received input frames to produce optimal results under low-light conditions. The method may include, by the transformation module 214, updating the parameter fractions using changes in image capture conditions and controlling the parameter fractions based on metadata.

[0053] Subsequently, method 300 may include performing multi-frame blending on the artifact-corrected input frame at operation 313. This operation may be performed by multi-frame blending module 216.

[0054] Figure 4A flowchart 400 illustrating an exemplary embodiment depicting a multi-frame blending scene according to embodiments of this subject matter is shown. In the embodiment, image sensor 210 receives multiple input image frames for multi-frame blending. The input frame images may be represented as 1, 2, 3, 4, 5, 6, 7. This operation may correspond to operation 301. Input frame analysis module 212 may be configured to determine at least one feature score extracted from the received multiple input frames. The multiple feature scores may include at least one of scores related to sharpness, noise, brightness, and light intensity scores. This operation may correspond to operation 303. Furthermore, input frame analysis module 212 may also be configured to determine a parameter score for each frame based on input features. In some embodiments, input frame analysis module 212 may be configured to determine the parameter score for each frame based on input features and / or metadata of the input frames. This operation may correspond to operation 305. The parameter score may be represented as PS = f(SS, NS, BS, ..., N). Furthermore, transformation module 214 may be configured to apply preprocessing corresponding to the intensity of each of the multiple input frames based on the parameter score. Preprocessing may include at least one of heavy or light denoising, heavy or light sharpening. For example, denoising of frames 2, 3, and 6 and sharpening of frames 1 and 2. This operation may correspond to operation 311. Subsequently, the multi-frame blending module 216 may be configured to blend one or more frames from multiple frames into a blended image output. This operation may correspond to operation 313.

[0055] Figure 5a and Figure 5b A flowchart illustrating an exemplary embodiment of weak light denoising using the proposed solution is shown, illustrating an embodiment of this subject matter. In the example, such as... Figure 5a As shown, an image sensor can be configured to receive bursts of raw, noisy images. The received image frames can then be preprocessed using existing techniques before multi-frame mixing. However, the mixed output may contain artifacts, including noise, blurred edges, and color artifacts. Therefore, existing techniques are not suitable for low-light capture. Furthermore, when processing the same received input image frames using the proposed solution, the mixed output image is remarkably sharp in extremely low-light denoising. Figure 5b As shown, the image sensor can be configured to receive bursts of raw, noisy images. The received image frames can then be processed by the input frame analysis module 212 and the transformation module 214. The resulting mixed output image can be clearer.

[0056] Figure 6a and Figure 6b A flowchart illustrating an exemplary embodiment of sharpness-based partial frame denoising, according to embodiments of this subject matter, is shown. Figure 6a The image shown is a composite image using existing technology, exhibiting noise, blurred edges, and color artifacts. Figure 6b The image obtained after denoising the input frames using the proposed solution is shown. In this embodiment, denoising is performed on 8 out of 10 input frames based on sharpness.

[0057] Figure 7a and Figure 7b A diagram illustrating an embodiment of this subject matter is shown, including another exemplary embodiment of sharpness-based partial frame denoising. Figure 7a A hybrid image output using existing technology is shown. The hybrid image output has noise and blurred edges. In an embodiment, Figure 7b The image obtained after denoising the input frame using the proposed solution is shown. Denoising of the input frame can be based on sharpness.

[0058] Figure 8a and Figure 8b A diagram illustrating yet another exemplary embodiment of sharpness-based partial frame denoising according to embodiments of this subject is shown. Figure 8a The image shown is a composite image created using existing techniques. Therefore, existing techniques are not suitable for low-light capture. Figure 8b The image obtained after denoising the input frame based on sharpness using the proposed solution is shown.

[0059] Figure 9a and Figure 9b The illustration depicts yet another exemplary embodiment of sharpness-based partial frame denoising using the proposed solution, illustrating an embodiment of this subject matter. Figure 9a The image shown is a composite image using existing technology, exhibiting noise, blurred edges, and color artifacts. Figure 9b The image obtained after denoising the input frame based on sharpness using the proposed solution is shown.

[0060] In view of the foregoing, various advantageous features relating to this disclosure are provided:

[0061] • Control over various output parameters, such as noise reduction level, edge enhancement, and brightness control, and

[0062] • Improved multi-frame blending output, especially in low-light and extremely low-light conditions.

[0063] While this disclosure has been described using specific language, it is not intended to create any limitation as a result. It will be apparent to those skilled in the art that various process modifications can be made to the described methods to achieve the inventive concepts taught herein. The accompanying drawings and the foregoing description provide examples of embodiments. Those skilled in the art will understand that one or more of the described elements can be well combined into a single functional element. Optionally, certain elements may be divided into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, the order of processes described herein may be changed and is not limited to the manner described herein.

[0064] Furthermore, the actions in any flowchart need not be performed in the order shown; nor is it necessary to execute all actions. Moreover, those actions that do not depend on other actions can be performed in parallel with other actions. The scope of the embodiments is by no means limited to these specific examples. Many variations (such as differences in structure, size, and material use) are possible, whether explicitly stated in the specification or not. The scope of the embodiments is at least as broad as that given by the appended claims.

[0065] The benefits, other advantages, and solutions to problems have been described above with respect to specific embodiments. However, the benefits, advantages, solutions to problems, and any components that may cause any benefit, advantage, or solution to occur or become more apparent should not be construed as key, necessary, or essential features or components of any or all claims.

Claims

1. A method for enhancing image quality, the method comprising: Receive multiple input frames and metadata from the image sensor; Determine one or more feature scores of the received input frames among the plurality of input frames; Based on the analysis of the determined one or more feature scores and the metadata of the received input frame, the parameter score of the received input frame is determined; Based on the parameter scores, one or more artifacts in the received input frame are identified for correction. Based on the parameter score, at least one of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), or Multiscale Structural Similarity Index (MS-SSIM) is obtained to assess perceived quality, and the quality of at least one of one or more features of the received input frame is estimated based on at least one of the PSNR, the SSIM, or the MS-SSIM assessed perceived quality. The one or more features are prioritized based on the estimated quality of the received input frame; Based on the estimated quality and the priority ranking, determine the required correction intensity for at least one artifact identified in the received input frame; The determined correction strength is applied to the received input frame; as well as Multi-frame mixing is performed on multiple received input frames for which the determined correction strength has been applied. Determining the parameter score includes: One or more vector scores of at least one feature score of the received input frame are generated based on a weighted average of the one or more feature scores of the received input frame. The one or more vector scores generated from the one or more feature scores of the received input frame are mixed; Scaling the mixed vector fractions; and Based on the scaling of the mixed vector scores, one or more feature vector scores are associated with each of one or more generated feature vectors.

2. The method according to claim 1, wherein, Determining the one or more feature scores includes: Context data is used to extract at least one feature of the received input frame, wherein the context data is determined based on one or more image capture conditions, including ISO, exposure time, and lighting conditions for each frame; and A score for the at least one extracted feature is calculated based on the level of the at least one extracted feature from the received input frame.

3. The method according to claim 1, further comprising: The parameter fractions of the received input frames are adjusted to produce optimal results under low-light conditions.

4. The method according to claim 1, further comprising: The parameter scores are updated using changes in image capture conditions; as well as The parameter scores are controlled based on the metadata.

5. The method according to claim 1, wherein, The metadata includes at least one of the ISO value, BV value, and exposure value of the image sensor.

6. A system for enhancing image quality, the system comprising: Image sensor; Input frame analysis module; Transformation module; as well as At least one processor, Wherein, the at least one processor is configured to: Receive multiple input frames and metadata from the image sensor; Determine one or more feature scores of the received input frames from the plurality of received input frames; The parameter score of the received input frame is determined based on the analysis of one or more feature scores and the metadata of the received input frame; Based on the parameter scores, one or more artifacts in the received input frame are identified for correction. The transformation module is configured as follows: Based on the parameter score, at least one of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), or Multiscale Structural Similarity Index (MS-SSIM) is obtained to assess perceived quality, and the quality of at least one of one or more features of the received input frame is estimated based on at least one of the PSNR, the SSIM, or the MS-SSIM assessed perceived quality. The one or more features are prioritized based on the estimated quality of the received input frame; Based on the estimated quality and the priority ranking, determine the required correction intensity for at least one artifact identified in the received input frame; The determined correction strength is applied to the received input frame; and Multi-frame mixing is performed on multiple received input frames for which the determined correction strength has been applied. The input frame analysis module for determining the parameter score by the at least one processor is configured as follows: One or more vector scores of at least one of the one or more feature scores of the received input frame are generated based on a weighted average of the one or more feature scores of the received input frame. The one or more vector scores generated from the one or more feature scores of the received input frame are mixed; Scaling the mixed vector fractions; and Based on the scaling of the mixed vector scores, one or more feature vector scores are associated with each of one or more generated feature vectors.

7. The system according to claim 6, wherein, The input frame analysis module for determining one or more feature scores by the at least one processor is configured as follows: At least one feature of the received input frame is extracted using context data, wherein the context data is determined based on one or more image capture conditions, including ISO, exposure time, and lighting conditions for each frame; and A score for the at least one extracted feature is calculated based on the level of the at least one extracted feature from the received input frame.

8. The system according to claim 6, wherein, The transformation module is configured to adjust the parameter fractions of the received input frames to produce optimal results under low-light conditions.

9. The system according to claim 6, wherein, The transformation module is also configured to update the parameter scores using changes in image capture conditions and to control the parameter scores based on the received metadata.

10. The system according to claim 6, wherein, The metadata includes at least one of the ISO value, BV value, and exposure value of the image sensor.

Images