Accuracy Of Target Recognition Research Articles

Predicting Performance Improvement of Human Activity Recognition Model by Additional Data Collection

The development of a machine-learning-based human activity recognition (HAR) system using body-worn sensors is mainly composed of three phases: data collection, model training, and evaluation. During data collection, the HAR developer collects labeled data from participants wearing inertial sensors. In the model training phase, the developer trains the HAR model on the collected training data. In the evaluation phase, the developer evaluates the trained HAR model on the collected test data. When the HAR model cannot achieve the target recognition accuracy, the developer iterates the above procedures by taking certain measures, including collecting additional training data, until the re-trained model achieves the target accuracy. However, collecting labeled data for HAR requires additional time and incurs high monetary costs. In addition, it is difficult to determine the amount and type of data to collect for achieving the target accuracy while reducing costs. To address this issue, this paper proposes a new method that predicts the performance improvement of the current HAR model, i.e., it determines the level of performance improvement achievable by re-training the HAR model with additional data, before collecting the additional data. Thus, the method enables the HAR developer to establish a strategy for additional data collection by providing advice such as "If labeled data for the Walking and Running activities from two additional participants is collected, the HAR accuracy of the current HAR model for Walking will improve by 20%." To achieve this, a neural network called AIP-Net is proposed to estimate the improvement in performance by analyzing the feature space of the current HAR model using the proposed entropy-based attention mechanism. The performance of AIP-Net was evaluated on eight HAR datasets using leave-one-dataset-out cross-validation.

CNN-Based Multiterrain Moving Target Recognition Model for Unattended Ground Sensor Systems

In recent years, many deep learning algorithms based on seismic signals have been proposed to solve the moving target recognition problem in unattended ground sensor systems. Despite the excellent performance of these deep networks, most of them can only be deployed on cloud-based devices and cannot be deployed on low-power hardware devices due to the large network size. Second, since seismic signals are affected by the terrain, employing only seismic signals as reconnaissance means for unattended ground sensors cannot achieve multiterrain-type adaptability. In response, this paper proposes an MFC-TinyNet method facing a multiterrain. The method adds depthwise separable convolutional layers to the network, which effectively reduces the size of the network while keeping the target recognition accuracy constant, and solves the problem that the model is difficult to deploy on low-power hardware. It also uses the Mel-frequency spectrum feature extraction method to fuse sound and seismic signals to improve the accuracy of the model’s moving target recognition on a multiterrain. Experiments demonstrate that the method can combine the two advantages of the small network model and multiterrain applicability.

Research on Video Target Detection and Tracking in Football Matches.

Computer vision is an interesting branch of artificial intelligence which is dedicated to how electronic devices can achieve the level of capabilities to perceive things just like ordinary human beings do. In order to solve the poor effect of video for the detection of target in football matches and the low accuracy of target tracking, this paper aims to make a deep exploration of the methods of video for the detection of target and tracking in football matches. The video moving for the detection of target method based on background model is used to extract the image in the background of the matching video which improves the light flow field. Secondly, the video differential image is acquired according to the difference of colors, the ghost target of the image in the video background model is scientifically determined, the ghost degree of the pixel points of the image is scientifically determined, and the flicker matrix of the target image is constructed. The number of pixels of the moving target is derived. A meanshift-based video target tracking algorithm is used in conjunction for the detection of target result to determine whether to track the target image until the overall video target tracking task is completed, move the central position of the target frame and background frame to the target position, select the best one to adapt to the target change, and determine whether to track the target image until the overall video target tracking task is completed. The simulation results suggest that the approach described in this study is capable of detecting and tracking moving objects, as well as improving target recognition and tracking accuracy.

Bubble Plume Target Detection Method of Multibeam Water Column Images Based on Bags of Visual Word Features

Bubble plumes, as main manifestations of seabed gas leakage, play an important role in the exploration of natural gas hydrate and other resources. Multibeam water column images have been widely used in detecting bubble plume targets in recent years because they can wholly record water column and seabed backscatter strengths. However, strong noises in multibeam water column images cause many issues in target detection, and traditional target detection methods are mainly used in optical images and are less efficient for noise-affected sonar images. To improve the detection accuracy of bubble plume targets in water column images, this study proposes a target detection method based on the bag of visual words (BOVW) features and support vector machine (SVM) classifier. First, the characteristics of bubble plume targets in water column images are analyzed, with the conclusion that the BOVW features can well express the gray scale, texture, and shape characteristics of bubble plumes. Second, the BOVW features are constructed following steps of point description extraction, description clustering, and feature encoding. Third, the quadratic SVM classifier is used for the recognition of target images. Finally, a procedure of bubble plume target detection in water column images is described. In the experiment using the measured data in the Strait of Georgia, the proposed method achieved 98.6% recognition accuracy of bubble plume targets in validation sets, and 91.7% correct detection rate of the targets in water column images. By comparison with other methods, the experimental results prove the validity and accuracy of the proposed method, and show potential applications of our method in the exploration and research on ocean resources.

Underwater fish image enhancement method based on color correction

Due to the absorption and scattering of light propagation underwater, the captured underwater images often have problems such as color bias, low contrast and poor clarity, resulting in low accuracy of underwater fish identification. To address this problem, this paper proposes an underwater fish image enhancement method based on color correction to enhance the acquired fish images and improve the accuracy of fish target recognition. Firstly, color correction is achieved by stretching the L component and changing the a and b components of the CIE-Lab color space to improve image sharpness. Then the colors of the R-G-B channels of the image are equalized to reduce the color bias. Finally, the histograms of the three R-G-B channels are redistributed. Comparison experiments were conducted with existing methods on a self-built fish image dataset, and the enhanced images were analyzed from both subjective and objective evaluations. The results showed that the enhancement effect of the method in this paper is better than other methods. Finally, the comparison experiments of target recognition before and after image enhancement were conducted on YOLOv5, and the results showed that the enhanced image target recognition accuracy was 99.8%, which was 1.2 percentage points higher than that before enhancement; the average accuracy mAP was 94.5%, which was 5.6 percentage points higher than that before image enhancement. The method in this paper can effectively improve the problems of underwater images and provide technical support for underwater target recognition.

A Measurement Image Translation-Automatic Target Recognition Technique Based on CycleGAN with SAR Simulation DB

The proposed approach achieves the reliable accuracy of synthetic aperture radar-automatic target recognition (SAR-ATR) with a simulation database. The simulation images of targets-of-interest are generated from inverse SAR using high-frequency techniques. A measurement image translation-automatic target recognition (MIT-ATR) uses two deep learning networks. The unique feature of the MIT-ATR is that the measurement images are translated to the simulation-like images by cycle generative adversarial network (CycleGAN). CycleGAN does not need to have a dataset of paired images between the measurement and simulation images. The generated simulation-like images are used as the inputs of the Visual Geometry Group (VGG) network. The VGG network is trained on a simulation database with a softmax layer of multi-classes. Five classes, including a T-72 tank, are considered in the numerical experiments. The images of each class are simulated at all azimuth angles, but the elevation angles range from 6° to 30°. The accuracy of the proposed approach is 63% better than that of the traditional method with only the VGG network. The simulation database could definitely supplement the lack of measurement data. The accuracy of MIT-ATR is properly handled by CycleGAN and the VGG network.

A Three-Dimensional Target Recognition Algorithm Utilizing Spatiotemporal Information Perception

In order to solve the low accuracy of three-dimensional target recognition by traditional algorithms, an improved three-dimensional target recognition model based on spatiotemporal convolutional neural network is established in this paper. The high-level feature presentation ability of deep neural network and the spatiotemporal perception performance of sequential modelling methods are combined perfectly in this model. Triplet loss function is also introduced to improve the representation of three-dimensional moving targets under different attitudes, which effectively reduces the feature distance of the same targets under various attitudes, and increases the feature distance of different targets at the same time. The performance of different feature extractors and aggregation algorithms is compared in this paper. Experimental results demonstrated that the ResNet is more suitable for feature extraction and the local soft attention is more effective in feature aggregation. Compared with the traditional three-dimensional target recognition methods, the spatiotemporal perception model proposed in this paper can achieve higher recognition accuracy for three-dimensional target recognition. At the same time, the calculation speed of the model is fast, which is conducive for its application.

Image Target Recognition Based on Multiregional Features under Hybrid Attention Mechanism

The growing volume of image data calls for better real-time performance of image feature extraction algorithms. To enhance the recognition accuracy of image targets, it is significant to build a more scientific deep learning network. Multimodal cross convolution or densely connected blocks have been introduced to classic deep learning networks, aiming to promote the recognition of image targets. However, these attempts fail to satisfactorily extract detailed features from the original image. To solve the problem, this paper explores the image target recognition based on multiregional features under hybrid attention mechanism. Specifically, a convolutional neural network (CNN) was established for extracting multiregional features based on the loss function of local feature aggregation. The model consists of three independent CNN modules, which are responsible for extracting the global multiregional features and the local features of different regions. Next, the channel domain attention mechanism and spatial domain attention mechanism were embedded in the proposed CNN, such that the model can recognize targets more accurately, without increasing the computing load. Finally, the proposed network was proved effective through the training and testing on a self-developed sample set of surveillance video images.

基于改进Mask-RCNN的玉米果穗损失检测方法

In order to quickly and accurately identify the corn ears lost during the corn harvesting process, a corn ear loss detection method based on the improved Mask-RCNN model was proposed. The lost corn ears in the field were taken as research objects, the images of the lost corn ears were collected and the fallen ears data set was established. The size ratio of the Anchor Box of the area recommendation network was changed by changing the K-means algorithm to reduce the influence of artificial setting intervention. The group convolution was introduced into the residual unit and the channel dimension was divided into 3 equal parts to reduce the model parameters in the basic feature extraction network ResNet. A Convolutional Block Attention Module (CBAM) was introduced to improve the accuracy of the model in the last layer of the ResNet network. Results showed that the average target recognition accuracy of the method on the test set in this study was 94.3%, which was better than that of the previous model, and the average time to recognize a single image was 0.320 s. The proposed method could detect the lost corn ears during the harvesting process under the complicated background, and provide a reference for the corn ear loss detection of the corn harvester.

Underwater acoustic target recognition method based on a joint neural network.

To improve the recognition accuracy of underwater acoustic targets by artificial neural network, this study presents a new recognition method that integrates a one-dimensional convolutional neural network and a long short-term memory network. This new network framework is constructed and applied to underwater acoustic target recognition for the first time. Ship acoustic data are used as input to evaluate the network performance. A visual analysis of the recognition results is performed. The results show that this method can realize the recognition and classification of underwater acoustic targets. Compared with a single neural network, the relevant indices, such as the recognition accuracy of the joint network are considerably higher. This provides a new direction for the application of deep learning in the field of underwater acoustic target recognition.

Amplification of the Fluorescence Signal with Clustered Regularly Interspaced Short Palindromic Repeats-Cas12a Based on Au Nanoparticle-DNAzyme Probe and On-Site Detection of Pb2+ Via the Photonic Crystal Chip.

Although great headway has been made in DNAzyme-based detection of Pb2+, its adaptability, sensitivity, and accessibility in complex media still need to be improved. For this, we introduce new ways to surmount these hurdles. First, a spherical nucleic acid (SNA) fluorescence probe (Au nanoparticles-DNAzyme probe) is utilized to specifically identify Pb2+ and its suitability for precise detection of Pb2+ in complex samples due to its excellent nuclease resistance. Second, the sensitivity of Pb2+ detection is greatly enhanced via the use of a clustered regularly interspaced short palindromic repeats-Cas12a with target recognition accuracy to amplify the fluorescent signal upon the trans cleavage of the SNA (signal probe), and the limit of detection reaches as low as 86 fM. Third, we boost the fluorescence on photonic crystal chips with a bionic periodic arrangement by employing a straightforward detection device (smartphone and portable UV lamp) to achieve on-site detection of Pb2+ with the limit of detection as low as 24 pM. Based on the abovementioned efforts, the modified Pb2+ fluorescence sensor has the advantages of higher sensitivity, better specificity, accessibility, less sample consumption, and so forth. Moreover, it can be applied to accurately detect Pb2+ in complex biological or environmental samples, which is of great promise for widespread applications.

Improving YOLOv4-Tiny’s Construction Machinery and Material Identification Method by Incorporating Attention Mechanism

To facilitate the development of intelligent unmanned loaders and improve the recognition accuracy of loaders in complex scenes, we propose a construction machinery and material target detection algorithm incorporating an attention mechanism (AM) to improve YOLOv4-Tiny. First, to ensure the robustness of the proposed algorithm, we adopt style migration and sliding window segmentation to increase the underlying dataset’s diversity. Second, to address the problem that YOLOv4-Tiny’s (the base network) framework only adopts a layer-by-layer connection form, which demonstrates an insufficient feature extraction ability, we adopt a multilayer cascaded residual module to deeply connect low- and high-level information. Finally, to filter redundant feature information and make the proposed algorithm focus more on important feature information, a channel AM is added to the base network to perform a secondary screening of feature information in the region of interest, which effectively improves the detection accuracy. In addition, to achieve small-scale object detection, a multiscale feature pyramid network structure is employed in the prediction module of the proposed algorithm to output two prediction networks with different scale sizes. The experimental results show that, compared with the traditional network structure, the proposed algorithm fully incorporates the advantages of residual networks and AM, which effectively improves its feature extraction ability and recognition accuracy of targets at different scales. The final proposed algorithm exhibits the features of high recognition accuracy and fast recognition speed, with mean average precision and detection speed reaching 96.82% and 134.4 fps, respectively.

Improved TLBO for Fusion of Infrared and Visible Images

Image fusion is an image enhancement method in modern artificial intelligence theory, which can reduce the pressure in data storage and obtain better image information. Due to different imaging principles, information of the infrared image and visible images’ information is complementary and redundant. The infrared image can be fused with a visible image to obtain both the high-resolution texture details and the edge contour of the infrared image. In this paper, the fusion algorithm of forest sample image is studied at the feature level, which aims to accurately extract tree features through information fusion, ensure data stability and reliability, and improve the accuracy of target recognition. The main research contents of this paper are as follows: (1) teaching learning-based optimization (TLBO) algorithm was used to optimize the weighted coefficient in the fusion process, and the value range of random parameters in the model was adjusted to optimize the fusion effect. Compared with before optimization, image information increased by 2.05%, and spatial activity increased by 15.27%. (2) Experimental data show that the target recognition accuracy of feature-level fusion results was 93.6%, 13.9% higher than that of the original infrared sample image, and 18.8% higher than that of the original visible sample image. Pixel-level and feature-level fusion have their characteristics and application scopes. This method can improve the quality of the specified region in the image and is suitable for detecting intelligent information in forest regions.

An Effective Approach for Sonar Image Recognition with Improved Efficientdet and Ensemble Learning

Abstract Precisely recognition of underwater targets plays a critical role in the field of underwater unmanned exploration. According to the characteristics of sonar image, in order to improve the accuracy of underwater automatic target recognition, this paper proposes an underwater target sonar image recognition method combined with improved Efficientdet and ensemble learning. According to the regional dominance principle and threshold segmentation method, the sonar image is preprocessed, and preliminary effective features extracted from preprocessed data by Efficientnet backbone, so as to balance the efficiency and power consumption of the algorithm. The feature fusion adopts the improved weighted bi-directional feature pyramid ( BiFPN ) structure to aggregate the global information and strengthen the feature representation of the shallow feature map, which optimizes the problem that the small feature extraction is not ideal due to the low resolution of the sonar image and less detail information. By building a two-level classifier to complete the learning task in an ensemble way, a better classification effect is obtained than a single classifier. This method is applied to the underwater intelligent perception group of the 10th National Marine Aircraft Design and Production Competition in 2021, and won the first prize in the country. The experimental results show that the designed ensemble network model is more accurate than the conventional convolution neural network in identifying underwater targets on the measured sonar image data set while ensuring a certain prediction speed.

Metric-based Few-shot Classification in Remote Sensing Image

Target recognition based on deep learning relies on a large quantity of samples, but in some specific remote sensing scenes, the samples are very rare. Currently, few-shot learning can obtain high-performance target classification models using only a few samples, but most researches are based on the natural scene. Therefore, this paper proposes a metric-based few-shot classification technology in remote sensing. First, we constructed a dataset (RSD-FSC) for few-shot classification in remote sensing, which contained 21 classes typical target sample slices of remote sensing images. Second, based on metric learning, a k-nearest neighbor classification network is proposed, to find multiple training samples similar to the testing target, and then the similarity between the testing target and multiple similar samples is calculated to classify the testing target. Finally, the 5-way 1-shot, 5-way 5-shot and 5-way 10-shot experiments are conducted to improve the generalization of the model on few-shot classification tasks. The experimental results show that for the newly emerged classes few-shot samples, when the number of training samples is 1, 5 and 10, the average accuracy of target recognition can reach 59.134%, 82.553% and 87.796%, respectively. It demonstrates that our proposed method can resolve fewshot classification in remote sensing image and perform better than other few-shot classification methods.

The Detection and Following of Human Legs Based on Feature Optimized HDBSCAN for Mobile Robot

Accurate target recognition and stable follow in real clutter environment are important tasks for following a random robot. In this paper, a scheme of detecting and tracking human legs using a single lidar is proposed. The HDBSCAN algorithm is optimized by human leg characteristics, and the leg line features are fused on the basis of the density characteristics of the leg laser radar data, and the leg information of the target is clustered and its location is obtained. Based on the relative shift angle of target moving and the lock-in algorithm of tracking distance, the single target detection and tracking in unstructured environment can be realized. The lidar dataset in multiple scenes verifies that the clustering algorithm based on feature optimization improves the accuracy of target legs detection by more than 10% compared with the original algorithm. The scheme performs follow-up experiments on a self-built solid robot. The experimental results show that the robot can achieve accurate and stable human target follow in clutter environment.

Optimization of College English Classroom Teaching Efficiency by Deep Learning SDD Algorithm.

In order to improve the teaching efficiency of English teachers in classroom teaching, the target detection algorithm in deep learning and the monitoring information from teachers are used, the target detection algorithm of deep learning Single Shot MultiBox Detector (SSD) is optimized, and the optimized Mobilenet-Single Shot MultiBox Detector (Mobilenet-SSD) is designed. After analyzing the Mobilenet-SSD algorithm, it is recognized that the algorithm has the shortcomings of large amount of basic network parameters and poor small target detection. The deficiencies are optimized in the following partThrough related experiments of student behaviour analysis, the average detection accuracy of the optimized algorithm reached 82.13%, and the detection speed reached 23.5 fps (frames per second). Through experiments, the algorithm has achieved 81.11% in detecting students' writing behaviour. This proves that the proposed algorithm has improved the accuracy of small target recognition without changing the operation speed of the traditional algorithm. The designed algorithm has more advantages in detection accuracy compared with previous detection algorithms. The optimized algorithm improves the detection efficiency of the algorithm, which is beneficial to provide modern technical support for English teachers to understand the learning status of students and has strong practical significance for improving the efficiency of English classroom teaching.

Bidirectional Denoising Autoencoders-Based Robust Representation Learning for Underwater Acoustic Target Signal Denoising

The marine environmental noise formed by wind noise, rain noise, biological noise, sea surface waves, seismic disturbances, and so on is a kind of interference background field in underwater acoustic channels, which brings adverse effects to underwater acoustic target recognition. To improve the recognition accuracy of underwater targets under background noise interference, a bidirectional denoising autoencoder (BDAE) is proposed in this paper for underwater acoustic target signal denoising robust representation learning. The proposed bidirectional denoising autoencoder is an extension of the regular denoising autoencoder, which uses the original underwater acoustic target signals and their corresponding denoised signals to learn robust representations. We then measure the usefulness of the learned representations using a support vector machine classifier. Our proposed approach is verified on the ShipsEar database. Experimental results indicate that the proposed bidirectional denoising autoencoder can effectively learn the robust representations of underwater acoustic target signal denoising and is superior to the traditional methods.

ОБЗОР МЕТОДОВ УЛУЧШЕНИЯ ВИЗУАЛЬНОГО КАЧЕСТВА ИЗОБРАЖЕНИЙ И ВИДЕО В НЕБЛАГОПРИЯТНЫХ ПОГОДНЫХ УСЛОВИЯХ

In the modern world, image processing is used in various fields of human activity. Due to various interference factors, image and video deterioration seriously reduces the accuracy and efficiency of target tracking and recognition. Hence, image restoration is becoming a major problem in the field of computer vision. Matrials and methods, results and discussions This study provides an overview of methods for improving the visual quality of images and videos when they are distorted by weather phenomena. The methods are classified according to the types and differences of weather phenomena. Examples of their approbation on images and video are given. Conclusions are drawn about each of the types of algorithms. Conclusion The article explored various methods to improve the visual quality of images and videos in adverse weather conditions. Each method was studied in detail, their advantages and disadvantages were considered, which made it possible to come to the following conclusions: 1) methods that use physical models are very efficient, but are computationally complex, in this regard, this method is more appropriate to use for image and video post-processing. 2) Methods based on histograms are simpler, but they are only suitable for static weather conditions (fog, haze, haze). 3) The most promising are methods based on learning. Neural networks allow solving more complex problems, due to the possibility of parallelizing information and further learning. Also, these methods do an excellent job of improving the quality of both images and videos. Learning-based methods are applicable to both static weather conditions (fog, haze, haze) and dynamic ones (snow, rain, hail), which makes them more versatile for solving this problem.

Low-SNR Recognition of UAV-to-Ground Targets Based on Micro-Doppler Signatures Using Deep Convolutional Denoising Encoders and Deep Residual Learning

The rapid development of flight control technology has made unmanned aerial vehicles (UAVs) widely used in high-precision strikes on the battlefield. The premise of this is to achieve accurate target recognition using UAV-based radars. Aiming at three typical ground targets, including pedestrians, wheeled vehicles, and tracked vehicles, the micro-Doppler modulation caused by the random vibration of the UAV is analyzed in this article for the first time. To improve the recognition accuracy under low signal-to-noise ratios (SNRs), Doppler signals are transformed into time–frequency images, and a deep convolutional denoising encoder (DCDE) is designed to effectively remove the noise without suppressing micro-Doppler characteristics. To avoid the complicated micro-Doppler feature extraction, deep residual learning that can reduce the burden of network training and gain higher learning efficiency compared with traditional deep convolutional neural networks (DCNNs) is adopted. Recognition results under various occasions using denoised micro-Doppler images and designed residual learning network indicate that the proposed method has higher precision and better robustness than current methods. Even when the SNR is only −16 dB, the overall recognition accuracy still exceeds 90%.