Image processing system based on incremental learning

Abstract

The present invention discloses an image processing system for incremental learning in the field of image processing, comprising the following steps: S1 is an image information input step, in which initial parameters are set for an incremental learning model in the system by inputting external image information into the system; S2 is a preprocessing step using an incremental learning algorithm, in which the input image information can be marked by preprocessing the image information, the incremental learning algorithm model classifies and determines the major content categories of the image information, and separately marks the information areas in the image, and sets the correlation between different information areas, and the preprocessing link can adjust the image color and saturation, thereby making the image part of the information area easier to identify, and the use of an incremental learning algorithm can combine multiple learning logics.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an image processing system for incremental learning. Background Art

[0002] Incremental learning is a machine learning paradigm that allows a model to learn gradually as it receives new data, rather than learning the entire dataset all at once. This learning approach is particularly useful for processing large datasets, real-time data streams, or constantly changing data environments. Here are some key characteristics and methods of incremental learning.

[0003] 1. Online Learning: In this mode, the model processes one data point or a small batch of data points at a time and updates its parameters based on these data points. This approach is suitable when there is a continuous stream of data.

[0004] 2. Continuous Learning: A challenge in incremental learning is avoiding catastrophic forgetting, where the model forgets knowledge from previous tasks while learning a new one. Continuous learning aims to mitigate this forgetting through various strategies, such as using memory replay, regularization techniques, or innovative neural network architectures.

[0005] 3. Lifelong Learning: This refers to the process by which a model continuously learns from multiple tasks or datasets throughout its lifetime. A lifelong learning model needs to be able to adapt to new tasks while retaining its knowledge of old tasks.

[0006] 4. Incremental training: In deep learning, incremental training allows a model to update its weights as it receives new data without retraining the entire network. This can be achieved through fine-tuning or using specific incremental learning algorithms.

[0007] 5. Experience Replay: This is a technique for storing previous experiences (e.g., by retaining a portion of the data or using a buffer) and reusing them when learning a new task to help the model retain its memory for old tasks.

[0008] 6. Elastic Weight Sharing: This is a regularization method that reduces catastrophic forgetting by penalizing weight changes that have a large impact on the performance of previous tasks.

[0009] 7. Progressive Neural Networks: This approach adapts to new tasks by gradually increasing the complexity of the network while maintaining the ability to learn old tasks.

[0010] 8. Combining multi-task learning with incremental learning: In some cases, multi-task learning and incremental learning can be used simultaneously to enable the model to learn multiple tasks simultaneously and adapt to the introduction of new tasks.

[0011] Incremental learning is very valuable in real-world applications, especially in environments where data is constantly updated and changing. However, it also brings some challenges, such as how to balance the learning of new and old knowledge, how to design model architectures that can adapt to new data, and how to effectively deal with catastrophic forgetting.

[0012] Traditional image processing algorithms have the following problems:

[0013] 1. Model training takes a long time;

[0014] 2. New products cannot be used directly;

[0015] 3. When new defects are discovered, a completely new model must be trained for testing. Summary of the Invention

[0016] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and in the abstract and title of the present invention to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0017] The purpose of the present invention is to address the technical problems existing in the background technology. The present invention proposes an incremental learning image processing system. The present invention can effectively improve the learning efficiency and recognition accuracy of image processing through this system. The incremental learning algorithm is used to combine and use multiple learning logics, so that the system can incrementally learn a recognition method to achieve the effect of multiple steps of image content recognition, and then improve the recognition accuracy of image content through multiple recognition logics. At the same time, the method of comparing multiple recognition logics with each other can achieve the effect of adaptive fine-tuning of recognition logic, thereby improving the recognition accuracy of image content after a period of use.

[0018] The present invention proposes an image processing system for incremental learning, comprising the following steps:

[0019] S1 is the step of inputting image information, which inputs external image information into the system to set initial parameters for the incremental learning algorithm model in the system;

[0020] S2 uses an incremental learning algorithm to perform preprocessing steps. By preprocessing image information, the input image information can be marked. The incremental learning algorithm model classifies and determines the major content categories of the image information, and separately marks the information areas in the image, and sets the correlation between different information areas. The preprocessing step can also adjust the image color and saturation, making the image in the information area easier to identify;

[0021] S3 uses a backtracking algorithm for analysis, inputting the pre-processed image information into the backtracking algorithm of the incremental learning algorithm for analysis. The backtracking algorithm summarizes past data and compares the input image information with previously analyzed image information, classifying the image information more accurately within the same category and identifying and marking the information content of different information areas.

[0022] S4 uses a fixed neural network algorithm for analysis. The pre-processed image information is input into the fixed neural network algorithm in the incremental learning algorithm for analysis. The fixed neural network algorithm uses a tree-like classification recognition method to define the image multiple times through tree-like level-by-level classification, and finally determines the image as the minimum level classification type. The fixed neural network algorithm performs independent level-by-level tree classification for different information areas, and finally integrates the results of multiple information area recognition to output the overall information content of the image information;

[0023] S5 compares the results of the two algorithms. This step compares and analyzes the recognition and classification results of the two algorithms for the same image information, thereby obtaining an algorithm that is more consistent with the preprocessing results obtained by the incremental learning algorithm. The two algorithms use a mutually independent recognition method to avoid the mutual influence of erroneous recognition results, so that erroneous recognition information will not be repeated;

[0024] S6 is a step of fine-tuning the incremental learning algorithm model. This step compares the results of the two algorithms with the pre-processing results obtained by the incremental learning algorithm, and fine-tunes the parameters of the algorithm with a large difference. In this way, the recognition logic unit within the algorithm is optimized by a reverse correction method to make its output result closer to the pre-processing result of the incremental learning algorithm.

[0025] S7 handles catastrophic forgetting. After the two algorithms have been fine-tuned, the image information is input into the preprocessing stage again. By re-preprocessing the same image, it can be determined whether the incremental learning algorithm model has deviated. At the same time, the fine-tuning information is recorded during the loop processing process, so that the algorithm can be quickly reset when the preprocessing changes, thereby solving the catastrophic forgetting problem.

[0026] S8 determines the incremental learning algorithm model steps. When the image information has been recognized for multiple cycles and the recognition results of the two algorithms for the image information are consistent and have a high accuracy rate, the current incremental learning algorithm model data is locked at this time, thereby determining the learning path from the input image information to the output of the recognition result of the internal content of the image information.

[0027] By adopting the above technical solution, this solution can achieve recognition learning efficiency and accuracy of image files through the incremental learning algorithm model. The incremental learning method can integrate multiple recognition logics with each other, and can add combinations based on new recognition methods discovered subsequently, thereby ensuring that the system is always in the latest recognition state.

[0028] Preferably, the preprocessing step is performed through an incremental learning algorithm, which preprocesses the image file. By analyzing the color distribution and information area distribution of the image file, the image file can be preliminarily marked. Since the preprocessing only analyzes the image file in a broad direction, the accuracy of the preprocessing result is relatively high, and the preprocessing result for the image file is more accurate.

[0029] By adopting the above technical solution, this solution can perform basic large-scale analysis and processing on image files through pre-processing processing effects, thereby providing the best reference for subsequent recognition while avoiding large deviations in analysis results.

[0030] Preferably, the analysis step using a backtracking algorithm has an independent image buffer. By storing the image files identified by the algorithm separately, the features of the image files previously identified by the algorithm can be compared and identified during the algorithm's working identification process, thereby achieving content tagging of the current input image file based on the labels marked with the past image files.

[0031] By adopting the above technical solution, this solution can analyze and identify current unknown pictures through past picture records, and recognition based on past data can effectively improve the recognition efficiency of pictures.

[0032] Preferably, the fixed neural network algorithm performs the analysis step, which will not be affected by the past image file library compared to the backtracking algorithm. The fixed tree-like classification and recognition method can gradually refine the internal classification results of the image file, and finally complete the minimum level classification of different information areas in the image file, and then achieve the overall recognition effect of the image file through multiple minimum level classification effects.

[0033] By adopting the above technical solution, this solution can gradually refine the image content by using a step-by-step recognition and classification method through a fixed structure of the neural network algorithm, thereby achieving the effect of independent recognition of the image. In contrast, the neural network algorithm will not be affected by past recognition data.

[0034] Preferably, the step of fine-tuning the incremental learning algorithm model can adaptively adjust the system during the process of identifying different types of image files by fine-tuning the algorithm in reverse, thereby improving the accuracy of targeted identification of different types of image files by multiple algorithms.

[0035] By adopting the above technical solution, this solution can determine the authenticity and accuracy of the current image content by comparing the results of the two algorithms, and evaluate two different and independent algorithms at the same time.

[0036] Preferably, the step of processing catastrophic forgetting backs up the fine-tuning data so that when a catastrophic deviation occurs in the incremental learning algorithm model, the algorithm parameters can be quickly restored according to the backup data, thereby avoiding the collapse of the overall learning algorithm.

[0037] By adopting the above technical solution, this solution can achieve the effect of self-correction of the recognition algorithm through cyclic iteration, so that the image recognition can be more accurate as the system is used.

[0038] In summary, the present invention includes at least one of the following beneficial effects: Through this system, the internal content of image files can be recognized, and image files can be marked according to the image content, thereby realizing the recognition and analysis of image content. This system adopts a combination of multiple recognition logics to improve the accuracy of image recognition, and can fine-tune the learning algorithm itself through adaptive loops, thereby achieving the effect of correcting the algorithm itself during the image recognition process, thereby improving the recognition accuracy of this system. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Figure 1 A front view of an embodiment of an image processing system for incremental learning according to the present invention; DETAILED DESCRIPTION

[0040] The following is combined with Figure 1 The present invention is described in further detail.

[0041] Example 1

[0042] like Figure 1 As shown, in this embodiment, in order to solve the existing problems, the present invention discloses an image processing system for incremental learning, including the following steps:

[0043] S1 is the step of inputting image information, which inputs external image information into the system to set initial parameters for the incremental learning algorithm model in the system;

[0044] S2 uses an incremental learning algorithm to perform preprocessing steps. By preprocessing image information, the input image information can be marked. The incremental learning algorithm model classifies and determines the major content categories of the image information, and separately marks the information areas in the image, and sets the correlation between different information areas. The preprocessing step can also adjust the image color and saturation, making the image in the information area easier to identify;

[0045] S3 uses a backtracking algorithm for analysis, inputting the pre-processed image information into the backtracking algorithm of the incremental learning algorithm for analysis. The backtracking algorithm summarizes past data and compares the input image information with previously analyzed image information, classifying the image information more accurately within the same category and identifying and marking the information content of different information areas.

[0046] S4 uses a fixed neural network algorithm for analysis. The pre-processed image information is input into the fixed neural network algorithm in the incremental learning algorithm for analysis. The fixed neural network algorithm uses a tree-like classification recognition method to define the image multiple times through tree-like level-by-level classification, and finally determines the image as the minimum level classification type. The fixed neural network algorithm performs independent level-by-level tree classification for different information areas, and finally integrates the results of multiple information area recognition to output the overall information content of the image information;

[0047] S5 compares the results of the two algorithms. This step compares and analyzes the recognition and classification results of the two algorithms for the same image information, thereby obtaining an algorithm that is more consistent with the preprocessing results obtained by the incremental learning algorithm. The two algorithms use a mutually independent recognition method to avoid the mutual influence of erroneous recognition results, so that erroneous recognition information will not be repeated;

[0048] S6 is a step of fine-tuning the incremental learning algorithm model. This step compares the results of the two algorithms with the pre-processing results obtained by the incremental learning algorithm, and fine-tunes the parameters of the algorithm with a large difference. In this way, the recognition logic unit within the algorithm is optimized by a reverse correction method to make its output result closer to the pre-processing result of the incremental learning algorithm.

[0049] S7 handles catastrophic forgetting. After the two algorithms have been fine-tuned, the image information is input into the preprocessing stage again. By re-preprocessing the same image, it can be determined whether the incremental learning algorithm model has deviated. At the same time, the fine-tuning information is recorded during the loop processing process, so that the algorithm can be quickly reset when the preprocessing changes, thereby solving the catastrophic forgetting problem.

[0050] S8 determines the steps of the incremental learning algorithm model. When the image information has been recognized for multiple cycles and the recognition results of the two algorithms for the image information are consistent and have a high accuracy rate, the current incremental learning algorithm model data is locked at this time to determine the learning path from the input image information to the output of the recognition results of the internal content of the image information.

[0051] Through the incremental learning algorithm, this system can combine multiple image recognition logics with each other and can reversely improve the recognition logic by comparing the recognition results, thereby realizing the adaptive image recognition ability of the device and playing the role of incremental learning.

[0052] Example 2

[0053] like Figure 1 As shown, in order to solve the existing problems, in this embodiment, based on the same concept as the above-mentioned embodiment 1, the incremental learning image processing system also includes: the preprocessing step of performing preprocessing through the incremental learning algorithm, which preprocesses the image file, and can preliminarily mark the image file through the color distribution and information area distribution of the image file. Since the preprocessing only analyzes the image file in a large category direction, the accuracy of the preprocessing result is relatively high, and the preprocessing result for the image file is more accurate.

[0054] The analysis step using a backtracking algorithm has an independent image buffer. By storing the image files identified by the algorithm separately, the features of the image files previously identified by the algorithm can be compared and identified during the algorithm's working identification process, thereby achieving content tagging of the current input image file based on the labels marked with the past image files.

[0055] The fixed neural network algorithm performs the analysis step, which is not affected by the past image file library compared to the backtracking algorithm. The fixed tree-like classification and recognition method can gradually refine the internal classification results of the image file, and finally complete the minimum level classification of different information areas in the image file, and then achieve the overall recognition effect of the image file through multiple minimum level classification effects.

[0056] The step of fine-tuning the incremental learning algorithm model can adaptively adjust the system during the process of identifying different types of image files by fine-tuning the algorithm in reverse, thereby improving the accuracy of targeted recognition of different types of image files by multiple algorithms.

[0057] The catastrophic forgetting processing step backs up the fine-tuning data so that when a catastrophic deviation occurs in the incremental learning algorithm model, the algorithm parameters can be quickly restored according to the backup data, thereby avoiding the collapse of the overall learning algorithm.

[0058] The use of a looping and iterative approach enables multiple recognitions of the same photo, thereby ensuring accurate recognition and fine-tuning the algorithm based on the accuracy of subsequent recognitions. The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes based on the structure, shape, and principle of the present invention should be included within the scope of protection of the present invention.

Claims

1. An image processing system for incremental learning, comprising the following steps: S1 is the step of inputting image information, which inputs external image information into the system to set initial parameters for the incremental learning algorithm model in the system; S2 uses an incremental learning algorithm to perform preprocessing steps. By preprocessing image information, the input image information can be marked. The incremental learning algorithm model classifies and determines the major content categories of the image information, and separately marks the information areas in the image, and sets the correlation between different information areas. The preprocessing step can also adjust the image color and saturation, making the image in the information area easier to identify; S3 uses a backtracking algorithm for analysis, inputting the pre-processed image information into the backtracking algorithm of the incremental learning algorithm for analysis. The backtracking algorithm summarizes past data and compares the input image information with previously analyzed image information, classifying the image information more accurately within the same category and identifying and marking the information content of different information areas. S4 uses a fixed neural network algorithm for analysis. The pre-processed image information is input into the fixed neural network algorithm in the incremental learning algorithm for analysis. The fixed neural network algorithm uses a tree-like classification recognition method to define the image multiple times through tree-like level-by-level classification, and finally determines the image as the minimum level classification type. The fixed neural network algorithm performs independent level-by-level tree classification for different information areas, and finally integrates the results of multiple information area recognition to output the overall information content of the image information; S5 compares the results of the two algorithms. This step compares and analyzes the recognition and classification results of the two algorithms for the same image information, thereby obtaining an algorithm that is more consistent with the preprocessing results obtained by the incremental learning algorithm. The two algorithms use a mutually independent recognition method to avoid the mutual influence of erroneous recognition results, so that erroneous recognition information will not be repeated; S6 is a step of fine-tuning the incremental learning algorithm model. This step compares the results of the two algorithms with the pre-processing results obtained by the incremental learning algorithm, and fine-tunes the parameters of the algorithm with a large difference. In this way, the recognition logic unit within the algorithm is optimized by a reverse correction method to make its output result closer to the pre-processing result of the incremental learning algorithm. S7 handles catastrophic forgetting. After the two algorithms have been fine-tuned, the image information is input into the preprocessing stage again. By re-preprocessing the same image, it can be determined whether the incremental learning algorithm model has deviated. At the same time, the fine-tuning information is recorded during the loop processing process, so that the algorithm can be quickly reset when the preprocessing changes, thereby solving the catastrophic forgetting problem. S8 determines the incremental learning algorithm model steps. When the image information has been recognized for multiple cycles and the recognition results of the two algorithms for the image information are consistent and have a high accuracy rate, the current incremental learning algorithm model data is locked at this time, thereby determining the learning path from the input image information to the output of the recognition result of the internal content of the image information.

2. The image processing system for incremental learning according to claim 1, characterized in that The preprocessing step is performed through an incremental learning algorithm, which preprocesses the image file. By analyzing the color distribution and information area distribution of the image file, the image file can be preliminarily marked. Since the preprocessing only analyzes the image file in a broad category direction, the accuracy of the preprocessing result is relatively high, and the preprocessing result for the image file is more accurate.

3. The image processing system for incremental learning according to claim 1, characterized in that The analysis step using a backtracking algorithm has an independent image buffer. By storing the image files identified by the algorithm separately, the features of the image files previously identified by the algorithm can be compared and identified during the algorithm's working identification process, thereby achieving content tagging of the current input image file based on the labels marked with the past image files.

4. The image processing system for incremental learning according to claim 1, characterized in that The fixed neural network algorithm performs the analysis step, which is not affected by the past image file library compared to the backtracking algorithm. The fixed tree-like classification and recognition method can gradually refine the internal classification results of the image file, and finally complete the minimum level classification of different information areas in the image file, and then achieve the overall recognition effect of the image file through multiple minimum level classification effects.

5. The image processing system for incremental learning according to claim 1, characterized in that The step of fine-tuning the incremental learning algorithm model can adaptively adjust the system during the process of identifying different types of image files by fine-tuning the algorithm in reverse, thereby improving the accuracy of targeted recognition of different types of image files by multiple algorithms.

6. The image processing system for incremental learning according to claim 1, characterized in that The catastrophic forgetting processing step backs up the fine-tuning data so that when a catastrophic deviation occurs in the incremental learning algorithm model, the algorithm parameters can be quickly restored according to the backup data, thereby avoiding the collapse of the overall learning algorithm.

Images