Systems and Soft Computing最新文献

The advent of computer technology and the modernization of sports education have led to an increasing reliance on intelligent technology in the field of sports education. To improve the application effect of intelligent technology in physical education, a volleyball motion analysis technology combined with pose estimation optimization algorithm is designed. During the process, Kinect collects the joint data of the research object, which is then combined with the commonly used background subtraction and frame difference optimization techniques. This results in the construction of a background model. The background update number in the initial frame is utilized as the reference value. The contour edge is smoothed through the application of a one-dimensional Gaussian kernel function, and a teaching action guidance system is designed. The experimental results showed that the average accuracy of the research method reached 79.7 % and the average recall rate reached 75.2 %. The average relative error of the method was 4.13 % when comparing the accuracy of human body model. The research method is validated to accurately capture and analyze volleyball motion, which can provide some technical help for sports teaching.

In order to improve English learning efficiency, this paper constructs a deep learning model of semantic orientation exploration based on English V+able corpus distribution and semantic roles. This article combines the practical needs of English learning and establishes an ILP model with the optimization objective of minimizing spectrum resource occupation. A traffic grooming based time aware multipath RSA algorithm (HMRSA-TG) is proposed to solve the standardization problem of English speech recognition. To improve the system efficiency of intelligent English learning systems, a traffic grooming based time aware multipath RSA algorithm (HMRSA-TG) is proposed. Through research, it has been verified that the semantic orientation exploration deep learning model based on the distribution of semantic roles in English V+able corpora can effectively improve the effectiveness of English speech learning. The corpus model proposed in this paper can provide a reliable benchmark database for many speech problem learners and play an important role in English translation software in recognizing input speech with different accents

Wind energy, as a renewable energy source, is becoming increasingly important. The maintenance and damage detection of wind turbine blades are particularly crucial. For this purpose, the study aims to optimize the You Only Look Once (YOLO) processing algorithm for drone images to improve the detection efficiency. Firstly, the damage images captured by drones are preprocessed and optimized, including deblurring, noise reduction, and image enhancement. Subsequently, the YOLOv5 model is improved in terms of structure and regression function, and a novel damage detection model is proposed. The research results indicated that the minimum loss function value of the improved model was 2.75, the average accuracy was 95 %, and the highest intersection over union was 91 %. After simulation testing, the detection effect of this model on abrasion, crackle, edge cracking, and coating peeling images was significantly better than other models in the same series. Its average time was as short as 2.43 s, reaching a maximum frame rate of 35.46. From this, the combination of drone image technology and improved image processing algorithm has a positive impact on improving the operational efficiency and safety of wind turbine blades. Compared with the traditional methods, the proposed model has significant advantages in terms of accuracy and real-time performance of damage detection, providing a new technical means for efficient maintenance of wind turbines. Meanwhile, the method shows high robustness and reliability in different types of damage detection, demonstrating the extensive potential in practical applications.

There are issues in current higher education long jump teaching, e.g., assessment relies on teachers' experience, lacks scientific evaluation, and can't quantitatively give performance feedback to students. To address these issues, this research first divides the long jump process into the approach run and mid-air phases. Secondly, it proposes a method for measuring approach run speed based on virtual line velocity algorithm. Subsequently, by combining the BlazePose human pose assessment algorithm with posture matching algorithms, a technique for assessing mid-air long jump movements integrated with BlazePose human pose assessment algorithm is designed. Finally, an intelligent long jump evaluation system incorporating the BlazePose human pose assessment algorithm is established. The research findings demonstrate that the average accuracy of video at 120FPS reaches a maximum of 94.47%. The assessment accuracy of mid-air long jump movements integrated with the BlazePose human pose assessment algorithm is highest, with accuracies of 94%, 90%, and 88% for the takeoff, hip extension, and abdominal contraction key movements respectively. Additionally, the method shows a scoring result with an average error range of 3 points compared to evaluations by professional teachers. In the practical application of the BlazePose human pose assessment algorithm's intelligent long jump evaluation system, evaluation scores and long jump proficiency receive scientifically objective assessments, while teachers provide targeted corrective feedback, achieving good application results. In summary, the proposed intelligent long jump evaluation system exhibits good performance, complete functionality, and can provide quantifiable data references for both teachers and students.

With the rapid development of science and technology, the field of sports is constantly exploring and applying new technical means to improve the training effect and competitive level of athletes. Among them, the athletes' posture detection technology based on the attitude solving algorithm has been widely concerned in recent years. However, the current attitude solving algorithm has the limitation of low precision and low efficiency. Aiming at this, a new attitude solving algorithm is proposed. Firstly, the coordinate system is determined according to the theory of inertial navigation, and the attitude Angle is obtained by calculating the acceleration and magnetic induction intensity. Then the current attitude matrix is calculated according to the obtained attitude Angle. The initializing quaternion based on the attitude matrix is studied. Then, according to the advantages and defects of the three sensors, a complementary filtering algorithm is proposed for data fusion, so as to reduce the error of the final attitude solution. In order to further improve the accuracy of attitude detection, the complementary filter algorithm and double-layer Kalman filter algorithm are combined to process the data, and finally the quaternion is updated. It can be seen that the detection error of the research constructed model is only 9.94%, and its three attitude angle errors are mainly concentrated between -0.5° and 0.5° The model constructed by the research can realize high-precision posture detection, which can provide more scientific and reliable training aids for gymnastics, which has very strict requirements for movements in sports. It has positive significance for the development of sports.

Leukemia is a type of cancer characterized by the exponential growth of abnormal blood cells, which damages white blood cells and disrupts the function of the human body’s bone marrow. It is very challenging to classify because blood smear images are complicated, and there is a lot of variation between each class. Acute Lymphoblastic Leukemia (B-ALL) is one of the subtypes of leukemia. It is a rapidly progressing cancer that originates in B lymphocytes, characterized by the overproduction of immature B lymphoblasts. The purpose of this work is to classify different types of B-ALL subtypes such as Benign, Malignant Early Pre-B, Malignant Pre-B, and Malignant Pro-B from the peripheral blood smear images effectively. To accomplish this task, a novel deep-learning technique based on a fine-tuned ResNet-50 model has been developed. Our fine-tuned ResNet-50 model integrates several additional customized fully connected layers, including dense and dropout layers. Various data augmentation techniques such as flipping, rotation, and zooming have been applied to mitigate the risk of overfitting. In addition, a five-fold cross-validation technique has been employed to enhance the model’s generalization. The performance of our proposed technique is compared with several other methods, including VGG-16, DenseNet-121, and EfficientNetB0, as well as existing baselines, using different performance metrics. Experimental results demonstrate the superiority of the fine-tuned ResNet-50 model, achieving the highest accuracy and an F1-score of 99.38%. It also outperforms existing state-of-the-art approaches by a significant margin. The proposed fine-tuned ReNet-50 model achieves such performance without the need for microscopic image segmentation which indicates its potential utility in healthcare sectors in enhancing precise leukemia diagnosis.

With the development of big data technology, the field of monitoring and improving teaching quality in universities has ushered in new opportunities and challenges. Big data technology enables the capture and analysis of massive amounts of data generated during the teaching process, providing the possibility for a deeper understanding of teaching activities. However, how to extract useful information from these vast amounts of data and transform it into strategies for teaching improvement is a challenge. The research aims to propose a teaching quality monitoring and improvement method based on big data technology, which combines K-means clustering algorithm and association rule mining algorithm to improve the accuracy of teaching monitoring and the effectiveness of teaching improvement. In order to cope with these challenges, the study proposes a research method of big data technology based on joint K-mean clustering algorithm and association rule mining algorithm. The study first analyzes the teaching quality monitoring and evaluation indexes using the K-mean algorithm. Then the association rule mining algorithm is utilized to mine the data in the teaching quality monitoring indicators with association rules on the basis of the obtained cluster analysis. Finally, on the basis of association rule mining, the study constructs the assessment model of teaching quality monitoring indicators by utilizing the fused method. The outcomes revealed that the average of data analysis accuracy and the average of recall rate of the modeling method were 93.79 % and 91.95 %, respectively. Meanwhile, the evaluation time of the modeling method in the process of teaching quality monitoring data processing was 17.3 s, and the evaluation precision was 93.15 % respectively. Additionally, the process's overall confidence and enhancement are 95.01 % and 86.73 %, respectively, and the modeling method's performance is compared to other approaches. This shown that the approach may significantly boost the precision and effectiveness of monitoring the quality of instruction, as well as offer strong backing for the enhancement of instruction in higher education institutions.

Automatic guided vehicles for logistics warehousing are a key link in the construction of intelligent logistics. To improve the positioning accuracy of warehouse robots, we designed an advanced extended Kalman filter method integrating multiple synchronous positioning techniques and map construction methods, and completed the calibration and detection of pallets based on color image information. The results revealed that the proposed multi-innovation enhanced model achieved minimum relative rotation and absolute trajectory errors of 0.13 and 0.09, outperforming existing models. It showcased excellent mapping fidelity and integrity (above 0.9) across various datasets, with a high loop detection success rate (0.91) enhancing map precision. The tray fusion detection algorithm's AUC (area under the curve) reached 0.92, reflecting a balanced accuracy-recall tradeoff. This research offers robust positioning and mapping capabilities in logistics warehousing environments, effectively identifying errors and ensuring pallet accuracy. The detection error and accuracy of this method are better than the other three models, with the lowest average absolute error of 0.32 and the lowest root-mean-square error of 0.27, and the overall error in the detection of pallets is small. The findings provide strong theoretical backing and technical support for advancing intelligent logistics warehousing technology. Precise positioning and identification capabilities enable logistics and warehousing robots to accurately and quickly complete tasks such as access, handling and sorting of goods, greatly improving the efficiency of warehousing operations, promoting the digital transformation and intelligent development of the logistics and warehousing industry, and improving the competitiveness of the industry and the level of service.

In construction management, the rationality of on-site layout is crucial for project progress, cost, and safety. In order to improve the rationality of on-site layout, a multi-objective optimization model combining ant colony algorithm and Pareto optimal solution was constructed based on genetic algorithm, and this model was applied to practical engineering cases. The results show that in terms of computational time, the genetic algorithm takes an average of 1702.0 s, while the improved algorithm takes an average of 421.0 s, which is 1281s less and 85.9% more than before the improvement. The performance of the improved algorithm is the best, and the optimal solution can be obtained through multiple iterations. The improved algorithm has improved the efficiency of on-site layout optimization, and possesses practical application value for the layout of construction management sites. It offers a certain reference for the reasonable setting of construction management sites.

This paper introduces a novel approach to image denoising that leverages the advantages of Generative Adversarial Networks (GANs). Specifically, we propose a model that combines elements of the Pix2Pix model and the Wasserstein GAN (WGAN) with Gradient Penalty (WGAN-GP). This hybrid framework seeks to capitalize on the denoising capabilities of conditional GANs, as demonstrated in the Pix2Pix model, while mitigating the need for an exhaustive search for optimal hyperparameters that could potentially ruin the stability of the learning process. In the proposed method, the GAN’s generator is employed to produce denoised images, harnessing the power of a conditional GAN for noise reduction. Simultaneously, the implementation of the Lipschitz continuity constraint during updates, as featured in WGAN-GP, aids in reducing susceptibility to mode collapse. This innovative design allows the proposed model to benefit from the strong points of both Pix2Pix and WGAN-GP, generating superior denoising results while ensuring training stability. Drawing on previous work on image-to-image translation and GAN stabilization techniques, the proposed research highlights the potential of GANs as a general-purpose solution for denoising. The paper details the development and testing of this model, showcasing its effectiveness through numerical experiments. The dataset was created by adding synthetic noise to clean images. Numerical results based on real-world dataset validation underscore the efficacy of this approach in image-denoising tasks, exhibiting significant enhancements over traditional techniques. Notably, the proposed model demonstrates strong generalization capabilities, performing effectively even when trained with synthetic noise.