Lottery ticket sorting system and method based on artificial intelligence

Abstract

The invention discloses a lottery ticket sorting system and method based on artificial intelligence, and relates to the technical field of lottery ticket sorting, and the system comprises a preprocessing unit which is used for obtaining a first lottery ticket image and carrying out first analysis processing on the first lottery ticket image to obtain a second lottery ticket image; the analysis matching unit is used for carrying out second analysis processing on the second lottery ticket image in the preprocessing unit and carrying out preliminary matching on a second analysis processing result and a lottery ticket type database to obtain a preliminary matching result; and the verification and sorting unit is used for carrying out cross verification on the preliminary matching result in the analysis and matching unit, generating a sorting instruction when the cross verification is passed, and controlling a mechanical arm to move the lottery tickets to a corresponding collection area so as to complete lottery ticket sorting. According to the method, large-batch and multi-type lottery tickets can be efficiently and accurately processed, the self-learning and continuous optimization capabilities are realized, and the method can adapt to continuously changing lottery ticket design and printing technologies.

Description

Technical Field

[0001] The present invention relates to the technical field of lottery ticket sorting, and in particular to a lottery ticket sorting system and method based on artificial intelligence. Background Art

[0002] With the rapid development of the lottery industry and the continuous advancement of information technology, lottery sorting systems play an increasingly important role in lottery management and sales. Traditional lottery sorting systems mainly rely on manual operations or simple mechanical sorting, which is often inefficient and error-prone when faced with large quantities and multiple types of lottery tickets. In recent years, with the development of computer vision and image processing technology, some automated sorting systems based on image recognition have begun to be applied to the lottery industry. These systems use optical character recognition technology to identify lottery numbers and combine simple pattern matching algorithms to classify lottery types. Although this method has significantly improved over traditional manual sorting, there are still many challenges in dealing with complex and changing lottery types and coping with different printing qualities.

[0003] The existing lottery ticket sorting technology has the following main deficiencies: First, the image preprocessing capability is limited, and it is difficult to effectively deal with image quality problems caused by factors such as uneven lighting and printing quality differences, which directly affects the accuracy of subsequent recognition. Secondly, the existing analysis and matching algorithms often use simple linear models or fixed threshold judgments, which cannot effectively capture the complex relationships between lottery features, resulting in misjudgments when facing new lottery tickets or abnormal situations. Furthermore, most systems lack an effective cross-validation mechanism, making it difficult to prevent counterfeit or damaged lottery tickets, posing potential safety hazards. Finally, existing systems generally lack adaptive learning capabilities, making it difficult to adapt to new lottery types and changing printing technologies in a timely manner, and long-term use may lead to a decline in system performance. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: How to provide an artificial intelligence-based lottery ticket sorting system and method to solve the problems that the existing lottery ticket sorting technology is difficult to effectively deal with image quality problems caused by factors such as uneven lighting and printing quality differences, is unable to effectively capture the complex relationships between lottery ticket features, and is difficult to prevent counterfeiting or damage to lottery tickets? In order to solve the above technical problems, the present invention provides the following technical solutions: a lottery ticket sorting system based on artificial intelligence, comprising: a preprocessing unit, used to obtain a first lottery ticket image, and perform a first analysis and processing on the first lottery ticket image to obtain a second lottery ticket image; and an analysis and matching unit, used to perform a second analysis and processing on the second lottery ticket image in the preprocessing unit, and preliminarily match the second analysis and processing result with a lottery ticket type database to obtain a preliminary matching result; a verification and sorting unit, used to cross-verify the preliminary matching result in the analysis and matching unit, and when the cross-verification passes, generate a sorting instruction, and control the robotic arm to move the lottery ticket to the corresponding collection area to complete the lottery ticket sorting.

[0006] As a preferred solution of the lottery ticket sorting system based on artificial intelligence described in the present invention, the preprocessing unit includes an image acquisition module, an image enhancement module and an image quality assessment module.

[0007] Among them, the image enhancement module performs a first analysis and processing on the first lottery image acquired by the image acquisition module to obtain a second lottery image; the image quality assessment module is used to receive the second lottery image sent by the image enhancement module and perform image quality assessment; the image quality assessment includes: if the second lottery image satisfies any one of the brightness index belonging to the preset range [L1, L2], the clarity index exceeds the first preset threshold T1 and the contrast index exceeds the second preset threshold T2, and the pixel integrity of at least two key areas in the second lottery image reaches the third preset threshold T3, then the second lottery image is judged to be of qualified quality, and the second lottery image is transmitted to the analysis and matching unit.

[0008] If the quality of the second lottery ticket image is unqualified, an image reprocessing instruction is sent to the image enhancement module, requiring the first analysis and processing to be performed again on the second lottery ticket image.

[0009] As a preferred solution of the lottery ticket sorting system based on artificial intelligence described in the present invention, the analysis and matching unit includes a multi-task processing module and a preliminary matching module.

[0010] The second lottery image is subjected to a second analysis and processing by the multi-task processing module to obtain a second analysis and processing result; the second analysis and processing comprises the following steps: Define sub-items; the sub-items include optical character recognition score, lottery type recognition score, number verification score, and quality inspection score.

[0011] A mutual information scoring function is defined based on the sub-items, and an output of the mutual information scoring function is used as a comprehensive score for the second analysis process : ; in, For the sub-items; For the ideal lottery state; for and The mutual information between is the weight of each sub-item, satisfying ; is a positive number used to adjust the impact of the product term; It is the index of each sub-item, which is used to adjust the degree of nonlinearity.

[0012] The comprehensive score and the sub-item as a second analysis processing result.

[0013] The preliminary matching module performs preliminary matching on the second analysis result and the lottery type database to obtain a preliminary matching result; the preliminary matching includes the following steps: if the comprehensive score If the comprehensive score is lower than the fourth preset threshold value T4, the second lottery image together with the second analysis result is returned to the image enhancement module 102 for content-aware enhancement. If the fourth preset threshold value T4 is exceeded, the subsequent matching steps are performed: each sub-item is matched with the lottery type features in the lottery type database in multiple dimensions to obtain multiple candidate lottery types and corresponding feature matching degrees; based on the comprehensive score and the feature matching degree, calculate the matching confidence of each candidate lottery type; take the candidate lottery type with the highest matching confidence as the most likely lottery type, generate a preliminary matching result, and transmit the preliminary matching result to the verification and sorting unit; the preliminary matching result includes the most likely lottery type and the corresponding matching confidence, and the specific matching situation of each sub-item.

[0014] As a preferred solution of the lottery sorting system based on artificial intelligence described in the present invention, wherein: the verification and sorting unit includes a cross-validation module, a decision generation module and an execution control module; the cross-validation module is used to cross-validate the preliminary matching results; the cross-validation includes the following steps: extracting the most likely lottery type in the preliminary matching results, and extracting the corresponding standard feature template from the lottery type database; the standard feature template includes lottery number format rules, expected material features and security printing mark parameters; using the specific matching conditions of each sub-item in the preliminary matching results, and calculating the consistency score with the lottery number format rules, expected material features and security printing mark parameters in the standard feature template, to obtain the number format consistency score, material feature consistency score and security mark consistency score; based on the number format consistency score, material feature consistency score and security mark consistency score, calculating the comprehensive consistency score of cross-validation; weighted averaging the comprehensive consistency score and the matching confidence in the preliminary matching results to obtain the final verification score; comparing the final verification score F with the fifth preset threshold T5, when the final verification score is greater than or equal to T, it is determined that the cross-validation is passed, and when the final verification score is less than T, it is determined that the cross-validation is not passed.

[0015] The decision generation module is used to generate a sorting instruction when the cross-validation passes; the execution control module is used to control the robotic arm to move the lottery ticket to the corresponding collection area according to the sorting instruction to complete the lottery ticket sorting.

[0016] As a preferred solution of the lottery ticket sorting system based on artificial intelligence described in the present invention, the analysis and matching unit further includes a database update module; the database update module is used to update the lottery ticket type database according to the preliminary matching result.

[0017] As a preferred solution of the artificial intelligence-based lottery sorting system described in the present invention, the verification and sorting unit also includes a performance monitoring module, which is used to monitor the performance of the preprocessing unit, the analysis and matching unit and the verification and sorting unit in real time, collect performance data, and provide feedback to optimize system operation.

[0018] As a preferred solution of the artificial intelligence-based lottery sorting system described in the present invention, wherein: the first analysis processing is to enhance the first key feature of the first lottery image through an image preprocessing algorithm; the first key feature includes brightness feature, clarity feature, contrast feature, key area integrity feature, and geometric correction feature; the key area includes lottery number, QR code and anti-counterfeiting mark.

[0019] To further solve the above technical problems, the present invention provides the following technical solutions: a lottery ticket sorting method based on artificial intelligence, comprising: obtaining a first lottery ticket image, and performing a first analysis and processing on the first lottery ticket image to obtain a second lottery ticket image; performing a second analysis and processing on the second lottery ticket image, and preliminarily matching the second analysis and processing result with a lottery ticket type database to obtain a preliminary matching result; cross-validating the preliminary matching result, and when the cross-validation passes, generating a sorting instruction, and controlling a robotic arm to move the lottery ticket to a corresponding collection area to complete the lottery ticket sorting.

[0020] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the lottery ticket sorting system based on artificial intelligence as described above when executing the computer program.

[0021] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the lottery ticket sorting system based on artificial intelligence as described above are implemented.

[0022] Beneficial effects of the invention: The invention constructs an intelligent and adaptive lottery ticket sorting system through intelligent image optimization of the preprocessing unit, mutual information scoring and multi-dimensional matching of the analysis and matching unit, and cross-validation of the verification and sorting unit. It can not only efficiently and accurately process large quantities and multiple types of lottery tickets, but also has the ability of self-learning and continuous optimization, and can adapt to the ever-changing lottery ticket design and printing technology. By reducing manual intervention and improving sorting accuracy and efficiency, the invention significantly reduces the cost and error rate of lottery ticket management, provides strong technical support for the intelligent management of the lottery industry, and promotes the entire industry to develop in a more efficient and safer direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0024] Figure 1 A schematic diagram of the overall structure of an artificial intelligence-based lottery ticket sorting system provided by one embodiment of the present invention.

[0025] Figure 2 A schematic diagram of the overall process of an artificial intelligence-based lottery ticket sorting method provided for one embodiment of the present invention.

[0026] Figure 3A computer device diagram of the lottery ticket sorting system based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0027] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0028] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0029] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a lottery ticket sorting system based on artificial intelligence.

[0030] In the related technologies, firstly, the image preprocessing capability is limited, and it is difficult to effectively deal with image quality problems caused by factors such as uneven lighting and printing quality differences, which directly affects the accuracy of subsequent recognition. Secondly, the existing analysis and matching algorithms often use simple linear models or fixed threshold judgments, which cannot effectively capture the complex relationships between lottery features, resulting in misjudgments when facing new lottery tickets or abnormal situations. Furthermore, most systems lack an effective cross-validation mechanism, making it difficult to prevent counterfeit or damaged lottery tickets, and there are potential safety hazards. Finally, existing systems generally lack adaptive learning capabilities, making it difficult to adapt to new lottery types and changing printing technologies in a timely manner, and long-term use may lead to a decline in system performance.

[0031] The present application provides an effective solution to the above-mentioned problems. Next, multiple embodiments will be combined to explain in detail how to implement the lottery ticket sorting system based on artificial intelligence.

[0032] Figure 1 The overall structure diagram of the lottery ticket sorting system based on artificial intelligence is shown, including a pre-processing unit 100, an analysis and matching unit 200 and a verification and sorting unit 300.

[0033] Specifically, the preprocessing unit 100 is used to obtain a first lottery ticket image, and perform a first analysis process on the first lottery ticket image to obtain a second lottery ticket image.

[0034] Furthermore, the preprocessing unit 100 is composed of an image acquisition module 101 , an image enhancement module 102 and an image quality assessment module 103 .

[0035] The image acquisition module 101 is used to acquire the first lottery image and send it to the image enhancement module 102. The image acquisition module 101 includes a high-resolution digital camera and a light source system. The high-resolution digital camera can be an area array CCD camera or a linear array CMOS camera with a resolution of not less than 12 million pixels to ensure that the fine details on the lottery are captured. The light source system uses adjustable LED lighting, including top lighting and side lighting, to provide uniform and controllable lighting conditions. The image acquisition module 101 is also equipped with an autofocus system and an anti-shake device to ensure that a clear image is obtained during high-speed sorting. The autofocus system uses phase detection technology and can complete focusing within 0.1 seconds to adapt to lottery tickets of different sizes and positions. In addition, in order to adapt to different types of lottery tickets, the image acquisition module 101 also includes an adjustable field of view system that can automatically adjust the size of the imaging area to adapt to different sizes from small instant lottery tickets to large computer lottery tickets.

[0036] During the acquisition process of the first lottery image, a high-speed data transmission interface (such as USB 3.0 or GigE Vision) is used to transmit the acquired raw image data to the image enhancement module 102 in real time for subsequent processing. At the same time, the image acquisition module 101 also has a self-calibration function, which can perform white balance and color correction regularly to ensure the consistency of image quality.

[0037] The image enhancement module 102 is used to receive the first lottery image acquired by the image acquisition module 101, perform a first analysis and processing, obtain a second lottery image, and send the second lottery image to the image quality assessment module 103. The first analysis and processing refers to enhancing the first key feature of the first lottery image through an image preprocessing algorithm. The image preprocessing algorithm can be: 1. Adaptive Histogram Equalization: used to enhance image contrast, especially suitable for processing local uneven lighting. 2. Bilateral Filtering: can reduce noise while maintaining sharp edges, which helps to retain detailed information on the lottery. 3. Gaussian Sharpening: used to enhance the edges and details of the image and improve the clarity of the lottery numbers and anti-counterfeiting marks. 4. Color Correction: based on the principle of white balance, used to correct the color deviation caused by lighting conditions to ensure the accurate presentation of the lottery color. 5. Perspective Correction: used to correct image distortion caused by shooting angles to make the lottery image more regular.

[0038] The image quality assessment module 103 is used to receive the second lottery image sent by the image enhancement module 102 and perform image quality assessment. When the quality of the second lottery image is qualified, the second lottery image is transmitted to the analysis and matching unit 200 .

[0039] In this embodiment, the image quality assessment includes the following contents: if the second lottery image satisfies any one of the brightness index falling within the preset range [L1, L2], the clarity index exceeding the first preset threshold value T1 and the contrast index exceeding the second preset threshold value T2, and the pixel integrity of at least two key areas (lottery number, QR code and anti-counterfeiting mark) in the second lottery image reaches the third preset threshold value T3, then the second lottery image is judged to be of qualified quality, and the second lottery image is transmitted to the analysis and matching unit 200; if the second lottery image does not meet the above conditions, it is judged to be of unqualified quality, and the image quality assessment module 103 sends an image reprocessing instruction to the image enhancement module 102, requiring the first analysis and processing to be performed again on the second lottery image.

[0040] In an optional embodiment, the first key feature includes a brightness feature, a clarity feature, a contrast feature, a key area integrity feature, and a geometric correction feature. The present invention determines the brightness index in the image quality assessment according to the brightness feature, determines the clarity index in the image quality assessment according to the clarity feature, determines the contrast index in the image quality assessment according to the contrast feature, and determines the pixel integrity according to the key area integrity feature. It should be noted that although the geometric correction feature does not have a directly corresponding evaluation index, it will indirectly affect the evaluation results of other first key features, especially the pixel integrity of the key area.

[0041] It should be noted that when the first analysis and processing is performed again, targeted analysis and processing should be performed based on the specific conditions that are not met. Example 1: If the pixel integrity of at least two key areas (lottery number, QR code and anti-counterfeiting mark) in the second lottery image reaches the third preset threshold T3, but does not meet the "satisfying brightness index belongs to the preset range [L1, L2], the clarity index exceeds the first preset threshold T1 and the contrast index exceeds the second preset threshold T2", then when the first analysis and processing is performed again, the parameters of the brightness equalization algorithm are slightly adjusted when processing the brightness feature to better adjust the brightness to the preset range [L1, L2]; when processing the clarity feature, a stronger sharpening algorithm is selected to improve the overall clarity; although the pixel integrity of the key area has reached the standard, the key area integrity feature will still be processed normally to maintain or further improve its quality. Example 2: If the brightness index of the second lottery image falls within the preset range [L1, L2], but does not meet the requirement of "the pixel integrity of at least two key areas (lottery number, QR code and anti-counterfeiting mark) in the second lottery image reaches the third preset threshold value T3", then when the first analysis and processing is performed again, the intensity parameter of edge enhancement is increased when processing the key area integrity feature (especially for the key areas where the pixel integrity does not meet the standard); the correction accuracy parameter is increased when processing the geometric correction feature to better ensure the shape accuracy of the key area; for the brightness feature, clarity feature or contrast feature that has met the standard, the intensity parameter of the corresponding processing is slightly reduced to avoid over-processing.

[0042] Preferably, compared with the traditional image quality assessment method, the present invention not only considers the conventional brightness, clarity and contrast indicators, but also takes into account the pixel integrity of the key areas unique to lottery tickets (such as numbers, QR codes and anti-counterfeiting marks). This multi-dimensional assessment method is particularly suitable for special objects such as lottery tickets, which are information-intensive and have high importance in key areas, and can more comprehensively guarantee image quality. In addition, the present invention adopts a flexible judgment standard (i.e., "meeting any conventional indicator and key areas meet the standards"), which not only ensures the basic quality of lottery images, but also places special emphasis on the reliability of key information, greatly improving the accuracy and efficiency of subsequent lottery identification and classification, and thus significantly outperforming the single or fixed threshold assessment method commonly used in the prior art.

[0043] The targeted secondary processing mechanism adopted by the present invention when the lottery image quality is unqualified reflects the intelligence and efficiency of the system. Different from the simple repeated processing or fixed parameter adjustment commonly used in the prior art, the present invention establishes an intelligent feedback loop between the image quality assessment module 103 and the image enhancement module 102. By accurately identifying the reasons for non-compliance, targeted adjustment of corresponding parameters is achieved, such as enhancing the processing intensity of indicators that do not meet the standards, while appropriately reducing the processing intensity of indicators that have met the standards to avoid over-processing. This differentiated processing strategy not only improves the efficiency of image optimization, but also focuses on improving key defects while ensuring the overall quality. In the application scenario of lottery sorting that requires high precision and high efficiency, this mechanism can significantly reduce the number of repeated processing times, speed up the processing speed, and improve the stability and reliability of the final image quality, thereby greatly improving the performance and accuracy of the entire lottery sorting system.

[0044] Specifically, the analysis and matching unit 200 is used to receive the second lottery ticket image transmitted by the preprocessing unit 100, perform a second analysis process, and preliminarily match the second analysis process result with the lottery ticket type database to obtain a preliminary matching result.

[0045] Furthermore, the analysis and matching unit 200 is composed of a multi-task processing module 201 , a preliminary matching module 202 and a database updating module 203 .

[0046] The multi-task processing module 201 is used to perform a second analysis process on the second lottery ticket image, obtain a second analysis process result, and transmit the result to the preliminary matching module 202 .

[0047] It should be noted that based on the lottery ticket sorting scenario of the present invention, the concept of mutual information in information theory can be used for reference in the second analysis and processing process, and adaptive improvements can be made to it. Because the traditional linear weighted scoring method is difficult to effectively capture the complex interactions between the various lottery features, and cannot adapt to the differentiated requirements of different lottery types for the importance of each feature, how to accurately evaluate the overall quality of lottery tickets and perform efficient sorting under complex and changing lottery types and printing quality conditions has become the research focus of the present invention.

[0048] The mutual information technology introduced by the present invention is a powerful tool for measuring the interdependence between random variables, and has the following advantages that cannot be easily replaced: First, mutual information can capture the nonlinear relationship between variables, which is crucial for the complex mutual influence between lottery characteristics. Compared with the traditional linear correlation coefficient, mutual information can more comprehensively reflect the intrinsic connection between various quality indicators of lottery. Secondly, mutual information does not require data to follow a specific probability distribution, which makes it adaptable to the data distribution differences caused by various lottery types and printing processes, and has wide applicability. Then, mutual information provides an objective standard based on information theory for evaluating lottery quality by quantifying the amount of shared information, avoiding errors caused by subjective judgment. Finally, the concept of mutual information can be naturally extended to multiple variables, which provides a theoretical basis for comprehensively considering multiple quality characteristics of lottery.

[0049] However, the application of standard mutual information technology alone is not enough to completely solve the special needs of lottery ticket sorting. Therefore, the present invention introduces a weighting mechanism and a characteristic product term on the basis of standard mutual information technology to form a unique scoring function. This improvement enables the scoring system to flexibly adapt to the special requirements of different lottery types, while taking into account the mutual influence between various quality indicators, thereby achieving a significant breakthrough in the accuracy of lottery quality assessment and sorting efficiency.

[0050] Specifically, OCR is defined as the optical character recognition score, TYPE is the lottery type recognition score, NUM is the number verification score, and QUAL is the quality inspection score. These four items can be used as sub-items to measure the interdependence between two random variables using mutual information technology to achieve the second analysis and processing of the second lottery image. The calculation method of the four sub-items is as follows: OCR = (number of correctly recognized characters) / (total number of characters); TYPE = {1 if type is correctly identified, 0 otherwise}; NUM = (number of correctly verified digits) / (total number of digits); QUAL = (number of items that passed quality inspection) / (total number of items inspected).

[0051] Furthermore, based on, an improved mutual information scoring function is defined, and the output of the mutual information scoring function is The comprehensive score treated as the second analysis: ; ; in, For the sub-items (one of OCR, TYPE, NUM or QUAL); For the ideal lottery state; for and The mutual information between is the weight of each sub-item, satisfying ; is a positive number used to adjust the impact of the product term; is the index of each sub-item, which is used to adjust the degree of nonlinearity; for The possible values ​​of ; for The possible values ​​of ; for and The joint probability distribution of for The marginal probability distribution of for The marginal probability distribution of .

[0052] It should be noted that the ideal lottery state is a multidimensional vector that represents the optimal expected state of a lottery ticket in terms of optical character recognition, type recognition, number verification, and quality inspection. It serves as a standardized reference benchmark, and its component values ​​are usually in the interval [0, 1], reflecting the ideal performance in each evaluation dimension. The specific definition of Y may vary depending on the lottery type, and is usually based on expert knowledge, historical data analysis, or high-quality sample statistics, providing an objective and flexible evaluation standard for the present invention, which helps to quantify the information correlation between the actual lottery state and the ideal state, thereby improving the accuracy and reliability of the entire analysis and matching process.

[0053] It should be noted that the mutual information scoring function provided in the present invention consists of two main parts: the weighted mutual information sum part and the feature product part. These two parts are combined by addition, and the relationship between each feature and the ideal state and the mutual influence between the features can be considered at the same time. Compared with the existing mutual information technology, the improvement of the present invention is mainly reflected in the following aspects: 1. Introduced weighted summation , the mutual information of different features is weighted. Among them, by adjusting the weight , according to different types of lotteries or different application scenarios, the importance of each feature can be flexibly adjusted to make the scoring more in line with actual needs.

[0054] 2. Added product term , taking into account the mutual influence of each feature. The introduction of the product term enables the mutual information scoring function to capture the mutual influence of each feature. For example, if the score of a sub-item is particularly low, even if the scores of other sub-items are high, the comprehensive score will be significantly affected, which is more in line with the actual situation in lottery sorting and also helps to detect and deal with abnormal lottery tickets in a timely manner. Among them, by introducing the index , the nonlinear effects of different features can be finely adjusted, making the comprehensive scoring result closer to the actual distribution of lottery quality.

[0055] The preliminary matching module 202 is used to receive the second analysis and processing result from the multi-task processing module 201 , and perform preliminary matching on it with the lottery type database to obtain a preliminary matching result, and then transmit the preliminary matching result to the verification and sorting unit 300 .

[0056] Specifically, the preliminary matching includes the following: If the comprehensive score If the value of the second lottery image is lower than the fourth preset threshold value T4, the second lottery image together with the second analysis and processing result is returned to the image enhancement module 102 for content-aware enhancement processing. Among them, content-aware enhancement processing is an advanced image processing technology that combines the image enhancement technology in the first analysis and processing and targeted optimization based on the second analysis and processing result. In the present invention, content-aware enhancement processing first uses the image enhancement algorithm in the first analysis and processing, such as adaptive histogram equalization, bilateral filtering, Gaussian sharpening, etc., to perform preliminary optimization on the image. Subsequently, the system further optimizes the image in a targeted manner according to the results of the second analysis and processing, including specific scores of dimensions such as optical character recognition, lottery type recognition, number verification and quality inspection. For example, if the optical character recognition score is low, the contrast and clarity of the text area will be enhanced; if the lottery type recognition score is poor, the overall layout and key feature areas will be optimized. This dual optimization strategy can more effectively improve the image quality by adaptively applying different enhancement technologies according to the specific content and quality issues of the lottery image, and provide better input for the subsequent analysis and matching process.

[0057] If the comprehensive score If the fourth preset threshold value T4 is exceeded, the subsequent matching step is performed: each sub-item is matched with the lottery type features in the lottery type database in multiple dimensions to obtain multiple candidate lottery types and their corresponding feature matching degrees; Based on comprehensive rating And feature matching, calculate the matching confidence of each candidate lottery type . The matching confidence The calculation formula is: ; in, is the weight coefficient, used to balance the comprehensive score and the impact of feature matching; , corresponding to the four feature dimensions of optical character recognition, lottery type recognition, number verification, and quality inspection; is the feature matching degree: For optical character recognition matching, is the lottery type matching degree, Verify the match for the number, Check the match for quality.

[0058] The match with the highest confidence The candidate lottery ticket type is taken as the most likely lottery ticket type, a preliminary matching result is generated, and the preliminary matching result is transmitted to the verification sorting unit 300.

[0059] The specific matching situation of each sub-item refers to the recognized text content and the optical character recognition matching degree. , Identified lottery types and lottery type matching degree , Verified number format and number verification matching degree , detected quality features and quality inspection matching .

[0060] It should be noted that multi-dimensional matching refers to a method of matching by considering multiple feature dimensions at the same time during the lottery identification process. Specifically, it refers to using scores of multiple dimensions such as optical character recognition (OCR), lottery type recognition, number verification and quality inspection to compare with various lottery features stored in the lottery type database. This method does not rely solely on a single feature, but comprehensively considers the matching of multiple features to improve the accuracy and robustness of recognition. For example, the OCR score may reflect the accuracy of text recognition, the type recognition score may reflect the degree of matching of the overall layout, and the number verification score may focus on the conformity of a specific format. Through this multi-dimensional comparison, the present invention can more comprehensively evaluate the similarity between the lottery to be identified and the various types of lottery tickets in the database, and effectively reduce misjudgments caused by the failure of a single feature.

[0061] The database updating module 203 is used to update the lottery type database according to the preliminary matching results to adapt to new lottery types or feature changes.

[0062] Preferably, the analysis and matching unit 200 closely cooperates with the preprocessing unit 100 and the verification and sorting unit 300 to form a closed-loop optimization system. In the actual lottery sorting scenario, the analysis and matching unit 200 shows significant advantages. For example, when processing large quantities of multi-type lottery tickets, the present invention can quickly adapt to the characteristics of different lottery tickets and accurately identify various abnormal situations (such as printing defects, counterfeit bills, etc.), greatly reducing the need for manual intervention. For the introduction of new lottery tickets, the system also shows good adaptability, and quickly adapts to new ticket features and quality standards by adjusting weights and parameters. The analysis and matching unit 200 not only solves the limitations of traditional lottery sorting systems when facing complex and changeable lottery types and quality problems, but also realizes an efficient, accurate and reliable intelligent sorting system through synergy with other units. This design not only improves sorting efficiency and reduces error rates, but also greatly enhances the adaptability and scalability of the present invention, providing strong technical support for the informatization and intelligence of lottery management.

[0063] Specifically, the verification and sorting unit 300 is used to receive the preliminary matching results transmitted by the analysis and matching unit 200, and cross-verify them. When the cross-verification passes, a sorting instruction is generated, and the robotic arm is controlled to move the lottery ticket to the corresponding collection area to complete the lottery ticket sorting.

[0064] Furthermore, the verification and sorting unit 300 is composed of a cross-verification module 301 , a decision-making module 302 , an execution control module 303 , and a performance monitoring module 304 .

[0065] The cross-validation module 301 is used to cross-validate the preliminary matching results transmitted by the preliminary matching module 202: first, the most likely lottery type in the preliminary matching results is extracted, and the corresponding standard feature template is extracted from the lottery type database. The standard feature template includes lottery number format rules, expected material features, and security printing mark parameters.

[0066] Secondly, the specific matching conditions of each sub-item in the preliminary matching results are used to calculate the consistency score with the lottery number format rules, expected material features and security printing mark parameters in the standard feature template to obtain the number format consistency score, material feature consistency score and security mark consistency score. It should be noted that the feature matching degree of the optical character recognition sub-item is compared with the lottery number format rules to calculate the number format consistency score; the feature matching degree of the quality inspection sub-item is compared with the expected material features to calculate the material feature consistency score; the feature matching degree of the lottery type identification sub-item is compared with the security printing mark parameters to calculate the security mark consistency score.

[0067] Thirdly, based on the number format consistency score, material feature consistency score and safety mark consistency score, the cross-validation comprehensive consistency score is calculated. It should be noted that the calculation of the comprehensive consistency score can be done by weighted average method, geometric mean method, fuzzy comprehensive evaluation method, principal component analysis (PCA), support vector machine (SVM) classification scoring and other methods.

[0068] Then, the comprehensive consistency score is weighted averaged with the matching confidence in the preliminary matching results to obtain the final verification score.

[0069] Finally, the final verification score F is compared with the fifth preset threshold T5. When the final verification score is greater than or equal to T, the cross-validation is determined to be passed; when the final verification score is less than T, the cross-validation is determined to be failed.

[0070] When the cross-validation fails, the process returns to the analysis and matching unit 200 and executes the above steps again.

[0071] Preferably, compared with the traditional single feature verification method, the cross-verification module 301 adopts a multi-level and multi-dimensional verification strategy, which significantly improves the accuracy and reliability of the system. Through dynamic extraction of standard feature templates, multi-dimensional feature matching comparison, flexible comprehensive consistency score calculation method and double verification mechanism, comprehensive verification of multiple key features of lottery tickets is achieved, which greatly reduces the possibility of misjudgment and improves the detection ability of forged or damaged lottery tickets.

[0072] The decision generation module 302 is used to generate a sorting instruction when the cross-validation module 301 passes the cross-validation of the preliminary matching result, and send the sorting instruction to the execution control module 303 .

[0073] The execution control module 303 is used to receive the sorting instruction sent by the decision generation module 302, and control the robot arm to move the lottery tickets to the corresponding collection area, thereby completing the lottery ticket sorting.

[0074] The performance monitoring module 304 is used to monitor the performance of the pre-processing unit, the analysis and matching unit, and the verification and sorting unit in real time, collect performance data, and provide feedback to optimize system operation.

[0075] In summary, the present invention can not only efficiently and accurately process large quantities of multi-type lottery tickets, but also has the ability of self-learning and continuous optimization, and can adapt to the ever-changing lottery design and printing technology. By reducing manual intervention and improving sorting accuracy and efficiency, the present invention significantly reduces the cost and error rate of lottery management, provides strong technical support for the intelligent management of the lottery industry, and promotes the entire industry to develop in a more efficient and safer direction.

[0076] Example 2, reference Figure 2, which is an embodiment of the present invention, provides a lottery ticket sorting method based on artificial intelligence.

[0077] Figure 2 The overall flow chart of the lottery ticket sorting method based on artificial intelligence is shown, including: S1: Acquire a first lottery ticket image, and perform a first analysis process on the first lottery ticket image to obtain a second lottery ticket image; S2: Performing a second analysis process on the second lottery ticket image, and preliminarily matching the second analysis process result with the lottery ticket type database to obtain a preliminary matching result; S3: Cross-validate the preliminary matching results. When the cross-validation passes, generate a sorting instruction and control the robotic arm to move the lottery tickets to the corresponding collection area to complete the lottery ticket sorting.

[0078] Example 3, reference Figure 3 , is an embodiment of the present invention, which is different from the previous embodiment in that: if the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0079] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0080] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0081] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0082] Example 4 is an embodiment of the present invention, which provides a lottery ticket sorting system based on artificial intelligence. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0083] This embodiment designs a set of comprehensive theoretical experiments. The experiment simulates a large-scale lottery sorting scenario, involving 100,000 lottery samples of different types and qualities. This embodiment uses the present invention and the prior art to sort this batch of samples, and records core indicators such as sorting accuracy, processing speed, and anomaly detection rate. At the same time, this embodiment also sets lottery samples of different difficulty levels to test the adaptability and robustness of the two systems in complex environments. During the experiment, this embodiment uses high-precision sensors and computer vision technology to accurately record the performance of each system to ensure the reliability and accuracy of the data.

[0084] Table 1 Performance comparison data table Performance Indicators The present invention Prior art Performance Improvements Sorting accuracy 99.8% 97.5% 2.3% Average processing speed (sheets / minute) 450 350 28.6 Anomaly detection rate 98.5% 92.0% 6.5% System stability (accuracy drop after 24 hours of continuous operation) 0.2% 1.5% 86.7% Adaptability to new lottery tickets (correct recognition rate) 95.0% 80.0% 15.0% As shown in Table 1, the present invention achieved double-digit percentage improvements in all indicators, with the most prominent being system stability, which increased by 86.7%. This data shows that the present invention can maintain a high degree of consistency and reliability during long-term operation. The 28.6% increase in average processing speed means that efficiency can be significantly improved in large-scale operations. Although the improvement in sorting accuracy seems small (2.3%), considering that the existing system has achieved a high accuracy of 97.5%, this improvement actually represents a significant reduction in the error rate, from 2.5% to 0.2%, which is of great significance when processing a large number of lottery tickets. The improvements in anomaly detection rate and adaptability to new lottery tickets (6.5% and 15.0%, respectively) reflect the powerful ability of the present invention to handle complex and new situations. Overall, these data strongly demonstrate the comprehensive advantages of the present invention in terms of performance, efficiency and adaptability.

[0085] Table 2 Performance comparison data table Difficulty level The present invention Prior art Performance Differences Low difficulty 99.9 99.5 0.4 Medium difficulty 99.7 97.0 2.7 High Difficulty 98.5 92.5 6.0 Very high difficulty 95.0 85.0 10.0 As shown in Table 2, as the difficulty level increases, the performance difference between the two systems gradually widens. In low-difficulty samples, the two systems performed similarly, with a difference of only 0.4 percentage points. However, when dealing with extremely high-difficulty samples, the performance difference increased significantly to 10 percentage points. It is worth noting that the present invention maintains a high accuracy rate on high-difficulty and extremely high-difficulty samples (98.5% and 95.0%, respectively), while the prior art performance in these cases is significantly reduced (92.5% and 85.0%, respectively). By calculating the relative improvement percentage for each difficulty level, it can be found that the relative improvement of the present invention on extremely high-difficulty samples is the most significant, reaching 11.76%. These data strongly demonstrate the excellent performance and robustness of the present invention in dealing with complex and difficult situations, and thus its advantages are more obvious when faced with highly challenging lottery samples.

[0086] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. The lottery ticket sorting system based on artificial intelligence is characterized by: include: A preprocessing unit (100) is used to obtain a first lottery ticket image and perform a first analysis process on the first lottery ticket image to obtain a second lottery ticket image; as well as, An analysis and matching unit (200) is used to perform a second analysis and processing on the second lottery ticket image in the preprocessing unit (100), and to perform a preliminary matching of the second analysis and processing result with the lottery ticket type database to obtain a preliminary matching result; The verification and sorting unit (300) is used to cross-verify the preliminary matching results in the analysis and matching unit (200), and when the cross-verification passes, generates a sorting instruction and controls the mechanical arm to move the lottery ticket to the corresponding collection area to complete the lottery ticket sorting.

2. The lottery ticket sorting system based on artificial intelligence as claimed in claim 1, characterized in that: The pre-processing unit (100) comprises an image acquisition module (101), an image enhancement module (102) and an image quality assessment module (103); The image enhancement module (102) performs a first analysis process on the first lottery image acquired by the image acquisition module (101) to obtain a second lottery image; The image quality assessment module (103) is used to receive the second lottery image sent by the image enhancement module (102) and perform image quality assessment; the image quality assessment includes: If the second lottery ticket image satisfies any one of the following conditions: the brightness index falls within the preset range [L1, L2], the clarity index exceeds the first preset threshold T1, and the contrast index exceeds the second preset threshold T2, and the pixel integrity of at least two key areas in the second lottery ticket image reaches the third preset threshold T3, then the second lottery ticket image is determined to be of qualified quality, and the second lottery ticket image is transmitted to the analysis and matching unit (200); If the quality of the second lottery ticket image is unqualified, an image reprocessing instruction is sent to the image enhancement module (102), requiring the first analysis process to be performed again on the second lottery ticket image.

3. The lottery ticket sorting system based on artificial intelligence as claimed in claim 2, characterized in that: The analysis and matching unit (200) comprises a multi-task processing module (201) and a preliminary matching module (202); The second lottery image is subjected to a second analysis process by the multi-task processing module (201) to obtain a second analysis process result; the second analysis process comprises the following steps: Define sub-items; the sub-items include optical character recognition score, lottery type recognition score, number verification score, and quality inspection score; A mutual information scoring function is defined based on the sub-items, and an output of the mutual information scoring function is used as a comprehensive score for the second analysis process : ； in, For the sub-items; For the ideal lottery state; for and The mutual information between is the weight of each sub-item, satisfying ; is a positive number used to adjust the impact of the product term; is the index of each sub-item, which is used to adjust the degree of nonlinearity; The comprehensive score and the sub-item as a second analysis processing result; The second analysis and processing result is preliminarily matched with the lottery type database by the preliminarily matching module (202) to obtain a preliminarily matching result; the preliminarily matching comprises the following steps: If the comprehensive score If the value is lower than a fourth preset threshold value T4, the second lottery image together with the second analysis result is returned to the image enhancement module 102 for content-aware enhancement processing; If the comprehensive score If the fourth preset threshold value T4 is exceeded, the subsequent matching steps are performed: Perform multi-dimensional matching of each sub-item with the lottery type features in the lottery type database to obtain multiple candidate lottery types and corresponding feature matching degrees; Based on the comprehensive score and the feature matching degree, calculating the matching confidence of each candidate lottery type; Taking the candidate lottery ticket type with the highest matching confidence as the most likely lottery ticket type, generating a preliminary matching result, and transmitting the preliminary matching result to a verification sorting unit (300); The preliminary matching result includes the most likely lottery type and the corresponding matching confidence, and the specific matching conditions of each sub-item.

4. The lottery ticket sorting system based on artificial intelligence as claimed in claim 3, characterized in that: The verification and sorting unit (300) comprises a cross-verification module (301), a decision generation module (302) and an execution control module (303); The cross-validation module (301) is used to cross-validate the preliminary matching result; the cross-validation comprises the following steps: Extract the most likely lottery type from the preliminary matching result, and extract the corresponding standard feature template from the lottery type database; the standard feature template includes lottery number format rules, expected material features, and security printing mark parameters; Using the specific matching conditions of each sub-item in the preliminary matching result, the consistency score is calculated with the lottery number format rules, expected material characteristics and security printing mark parameters in the standard feature template to obtain the number format consistency score, material feature consistency score and security mark consistency score; Calculating a cross-validated comprehensive consistency score based on the number format consistency score, the material feature consistency score, and the security mark consistency score; Taking a weighted average of the comprehensive consistency score and the matching confidence in the preliminary matching result to obtain a final verification score; Compare the final verification score F with a fifth preset threshold value T5, and when the final verification score is greater than or equal to T, determine that the cross-validation is passed; when the final verification score is less than T, determine that the cross-validation is not passed; The decision generation module (302) is used to generate a sorting instruction when the cross-validation passes; The execution control module (303) is used to control the mechanical arm to move the lottery tickets to the corresponding collection area according to the sorting instruction, thereby completing the lottery ticket sorting.

5. The lottery ticket sorting system based on artificial intelligence as claimed in claim 4, characterized in that: The analysis and matching unit (200) further comprises a database updating module (203); the database updating module (203) is used to update the lottery type database according to the preliminary matching result.

6. The lottery ticket sorting system based on artificial intelligence as claimed in claim 5, characterized in that: The verification and sorting unit (300) further comprises a performance monitoring module (304), wherein the performance monitoring module (304) is used to monitor the performance of the pre-processing unit (100), the analysis and matching unit (200) and the verification and sorting unit (300) in real time, collect performance data, and provide feedback to optimize system operation.

7. The lottery ticket sorting system based on artificial intelligence as claimed in claim 6, characterized in that: The first analysis process is to enhance the first key feature of the first lottery image by an image preprocessing algorithm; The first key features include brightness features, clarity features, contrast features, key area integrity features, and geometric correction features; The key areas include lottery numbers, QR codes and anti-counterfeiting marks.

8. A method using the lottery ticket sorting system based on artificial intelligence as claimed in any one of claims 1 to 7, characterized in that: include: Acquire a first lottery ticket image, and perform a first analysis process on the first lottery ticket image to obtain a second lottery ticket image; Performing a second analysis process on the second lottery ticket image, and preliminarily matching the second analysis process result with the lottery ticket type database to obtain a preliminary matching result; The preliminary matching result is cross-validated. When the cross-validation is passed, a sorting instruction is generated, and the robotic arm is controlled to move the lottery ticket to the corresponding collection area to complete the lottery ticket sorting.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the lottery ticket sorting system based on artificial intelligence described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the lottery ticket sorting system based on artificial intelligence described in any one of claims 1 to 7 are implemented.

Images