Pattern Recognition Algorithms Research Articles

Quantitative Evaluation of the Ability of Accelerated Tests to Simulate On-Road Corrosion Morphology of Automotive Aluminum Alloys

Three wrought automotive Al-Mg-Si alloys (6xxx-series) were exposed to an on-road service environment and in six standard laboratory-accelerated exposure tests. Optical micrographs of all these exposures were obtained and the resulting corrosion morphology was quantified using fractal dimension analysis and corrosion boundary length-to-area ratios. Additionally, the images were also used to train a convolutional neural network (CNN)-based pattern recognition algorithm, which was then used to quantitatively identify which accelerated test was the closest match to the field exposures. Overall, no single accelerated test could fully capture the on-road results or ubiquitously be the most appropriate test regardless of alloy and temper. However, results from fractal dimension and length-to-area ratio analyses identified that among the tests studied, those with acidified electrolytes are more appropriate for matching on-road corrosion morphology. The CNN algorithm output also agreed with this finding, indicating that the results from tests with acidified electrolytes correlated to field morphology with a confidence >70% for most of the images tested, thus showing the utility of these methods in providing quantitative bases for morphology comparison. Assessed in the context of literature evidence for localized corrosion mechanisms in 6xxx-series alloys, these results also indicated that pH may play an influential role in how corrosion morphology develops in these alloys upon exposure to a complex on-road environment.

High-performance and self-calibrating multi-gas sensor interface to trace multiple gas species with sub-ppm level

Developing a better understanding of gas-metal oxide interactions as well as improving the output performance with the aging of the gas sensing array represents a critical challenge. In this work, we have designed a surface-functionalized metal-oxide sensors array containing six units for detecting NO2, SO2, H2S, and CO hazardous gases. The responses of the sensor array to four target gases are enhanced by various catalytic effects on the surface, and they show distinctive responses to each gas. For improved selectivity, a pattern recognition algorithm is implemented under an edge computing-based environment to relieve the overload of a monitoring server. Moreover, a self-calibration method for long-term stability is designed to indirectly calibrate gas responses with aging. With its edge-computing device prototype, which includes a sensor array and its readout integrated circuit, four target gases are distinguished by an artificial neural network algorithm with 97% accuracy. The present work describes the effective platforms for emerging next-generation gas sensors with improved communication and monitoring systems.

Portable electronic nose system with elastic architecture and fault tolerance based on edge computing, ensemble learning, and sensor swarm

The portable electronic nose (E-nose) systems are suffering from the limited computing ability of microcontrollers and can only adopt simple pattern recognition algorithms. The heavy coupling between the sensors greatly limits the anti-fault ability of the system. Herein, a novel ensemble learning framework based on independent artificial neural networks (ANN) is proposed in a portable E-nose system to recognize volatile organic compounds (VOCs). Edge computing is applied to complete data processing and analysis on an ARM Cortex-M3 based microcontroller in an E-nose system. Each sensing unit can recognize VOCs by training independent ANN models. Then an ensemble learning method further organizes the diverse ANN models into an onboard sensor swarm with a significantly enhanced performance. The accuracy of the type classification can reach 81.1%, which is improved by more than 20% compared with the best individual classifier. The R2 score of concentration prediction can reach 84.1%, which is improved by more than 25% compared with the best regressor in the ensemble group. For both type identification and concentration prediction, the tolerance to the sensor faults of swarm-based E-nose is 10 and 18 times greater than that of the conventional array-based sensors, respectively. Our work has designed an elastic architecture based on sensor swarms, providing a novel avenue for the development of E-nose systems with high fault tolerance.

Gas Sensor Array with Pattern Recognition Algorithms for Highly Sensitive and Selective Discrimination of Trimethylamine

Artificial senses like electronic nose, which ameliorates the problem of poor selectivity from single gas sensor, have elicited keen research interest to monitor hazardous gases. Herein, the doping effects of gallium on In2O3 nanotubes (NTs) are investigated and a four‐component sensor array for the detection of trimethylamine (TMA) is reported. All‐gallium‐doped/alloyed In2O3 (Ga‐In2O3) sensors show improved sensitivity and selectivity to TMA at an operating temperature of 240 °C, with 5 mol% Ga‐doped/alloyed one displaying the highest response in the range of 0.5–100 ppm and the lowest detection limit of 13.83 ppb. Based on the gas‐sensing properties, a four‐component sensor array is fabricated, which shows unique response patterns in variable‐gas backgrounds. Herein, back propagation neural network (BPNN), radial basis function neural network (RBFNN), and principal component analysis‐based linear regression (PCA‐LR) are trained with the gas‐sensing data to discriminate different gases with high accuracy, as well as to predict the concentrations of target gases in different gases and gas mixtures. Furthermore, accuracies of 92.85% and 99.14% can be achieved for the classification of six gases (three single gases and three binary gas mixtures) and for the prediction of TMA concentrations in the presence of different concentrations of TMA and acetone, respectively.

Functional Impairment in Individuals Exposed to Violence Based on Electronical Forensic Medical Record Mining and Their Profile Identification: Controlled Observational Study.

Little is known about the functional consequences of violence when directly assessed as a primary outcome, and even less about how consistently these consequences are evaluated in a judicial context. The World Health Organization (WHO) highlighted the importance of a functional approach to health in 2001 with the release of the International Classification of Functioning, Disability, and Health (ICF). In most European countries, forensic physicians assess individuals exposed to violence to evaluate the outcomes of violence, providing certified medical evidence for magistrates' sentencing decisions. This evaluation involves a mix of objective, subjective, and contextual elements, such as reported symptoms of fear, pain, and details of the assault. Quantifying these subjective elements with scales could enhance their interpretation and application in a judicial context. This study aims to (1) characterize and (2) assess 6 scales measuring subjective elements of functional impairment among individuals exposed to violence. We conducted a retrospective study that included individuals exposed to violence examined in a French department of forensic medicine over 12 months. A typology of violence encountered in medical settings was built based on the mining of electronic health records and the use of pattern recognition algorithms. The optimal number of violence types was determined using a robust and stable clustering approach, involving sample resampling and a multimetric scheme. Patients were then paired according to their homogeneous profiles, and the intra- and interrater reproducibility of the scales was evaluated. All pain, fear, and life threat scales were significantly associated with higher functional impairment, suggesting that these measures contribute to the overall assessment of functional impairment. The intra- and interrater reproducibility of scales among similar situations of violence was measured, ranging from mild to good, with coefficients of concordance between 0.46-0.66 and 0.43-0.66, respectively. Individuals reporting intimate partner violence showed higher scores in both fear and perception of a life threat during the assault and medical interview, while individuals reporting battery by multiple unknown assailants presented higher scores only in perception of a life threat during the assault. We identified 5 remarkably stable profiles of situations of violence, consistent with clinical practice. Pain, fear, and life threat scales were related to functional impairment according to expert knowledge and demonstrated fair reproducibility under real-life conditions for similar situations of violence. Subjective elements related to functional impairment in individuals exposed to violence can be quantified using Likert scales during medical interviews.

Community engagement and data quality: best practices and lessons learned from a citizen science project on birdsong

Citizen Science (CS) is a research approach that has become popular in recent years and offers innovative potential for dialect research in ornithology. As the scepticism about CS data is still widespread, we analysed the development of a 3-year CS project based on the song of the Common Nightingale (Luscinia megarhynchos) to share best practices and lessons learned. We focused on the data scope, individual engagement, spatial distribution and species misidentifications from recordings generated before (2018, 2019) and during the COVID-19 outbreak (2020) with a smartphone using the ‘Naturblick’ app. The number of nightingale song recordings and individual engagement increased steadily and peaked in the season during the pandemic. 13,991 nightingale song recordings were generated by anonymous (64%) and non-anonymous participants (36%). As the project developed, the spatial distribution of recordings expanded (from Berlin based to nationwide). The rates of species misidentifications were low, decreased in the course of the project (10–1%) and were mainly affected by vocal similarities with other bird species. This study further showed that community engagement and data quality were not directly affected by dissemination activities, but that the former was influenced by external factors and the latter benefited from the app. We conclude that CS projects using smartphone apps with an integrated pattern recognition algorithm are well suited to support bioacoustic research in ornithology. Based on our findings, we recommend setting up CS projects over the long term to build an engaged community which generates high data quality for robust scientific conclusions.

Research on Embedded Ball Screw Diagnosis System for Mechanical Fault Based on Bispectral Image

Ball screw is one of the main basic transmission components in CNC machine tool feed system because of its high transmission efficiency, positioning accuracy, and rigidity. It plays an irreplaceable role in the process of machine tool torque transmission and positioning. However, ball screw is also one of the parts which are easy to break down in the feed system. If it breaks down, it will affect the processing accuracy of the whole machine tool and even cause shutdown. Numerical control means that digital control is a high-efficiency technology that uses digital information to control mechanical movement and processing processes. Therefore, to increase the high-speed and high-precision operation ability of CNC machine tools, research and development of efficient and reliable predictive maintenance and issue identification technology to guide maintenance of CNC machine tools are critical. Throughout this study, a bispectrum image-based embedded defect fight against the disease for cylindrical gears is provided. Firstly, the bispectrum of fault signal is generated by Wavelet Packet Threshold denoising, wavelet packet reconstruction, and bispectrum analysis, and the characteristics of bispectrum are represented by the first kind of gray moments. These feature statistics are used as input feature vectors of BP neural network pattern recognition algorithm to classify these feature sets and identify the corresponding gear fault types. Experiments show that under the condition of noise, after 14.5 hours of testing, the NGI has increased significantly and can be used to diagnose faults in the gear system, and the method achieves relatively ideal recognition rate and verifies the feasibility of gear box fault diagnosis method based on image recognition. The code compiled by OpenCV library is easy to transplant to embedded system, with high development efficiency and reliable program operation. OpenCV has the characteristics of powerful matrix computing capability and convenient and flexible interface, and it is a general algorithm that can realize real-time image processing.

A real-time fingerprint-based indoor positioning using deep learning and preceding states

In fingerprint-based positioning methods, the received signal strength (RSS) vectors from access points are measured at reference points and saved in a database. Then, this dataset is used for the training phase of a pattern recognition algorithm. Several noise types impact the signals in radio channels, and RSS values are corrupted correspondingly. These noises can be mitigated by averaging the RSS samples. In real-time applications, the users cannot wait to collect uncorrelated RSS samples to calculate their average in the online phase of the positioning process. In this paper, we propose a solution for this problem by leveraging the distribution of RSS samples in the offline phase and the preceding state of the user in the online phase.In the first step, we propose a fast and accurate positioning algorithm using a deep neural network (DNN) to learn the distribution of available RSS samples instead of averaging them at the offline phase. Then, the similarity of an online RSS sample to the RPs’ fingerprints is obtained to estimate the user’s location. Next, the proposed DNN model is combined with a novel state-based positioning method to more accurately estimate the user’s location. Extensive experiments on both benchmark and our collected datasets in two different scenarios (single RSS sample and many RSS samples for each user in the online phase) verify the superiority of the proposed algorithm compared with traditional regression algorithms such as deep neural network regression, Gaussian process regression, random forest, and weighted KNN.

Bio-inspired encoding for a real-time and stable single component odor detection with a highly-redundant optical artificial olfactory system

The standard model of artificial olfaction includes arrays of partially selective sensors and pattern recognition algorithms. Despite the brilliant results, such simplified platforms fail to achieve most of the natural olfaction features even if current sensors may rival olfactory receptor performances. Conversely, the olfaction architecture is specific to processing signals of millions of neurons, consisting of copious copies of a limited variety of receptors. The actual gas sensor arrays are still far from the abundance of natural olfaction receptors, so pattern recognition algorithms mainly differ from the processing in biological olfaction. This paper tries to cover this gap by applying a biologically derived processing algorithm to a highly redundant array. A webcam recorded the RGB channel color intensities of 15 optical indicators due to exposure to different chemical vapors in real time. Each pixel was assumed as a distinct sensor obtaining about 28,000 copies representative of 45 cross-selective virtual receptors. This outstanding redundancy enables deeply studying a model of the olfactory bulb configured as a network of inhibitory and excitatory elements. Subsequently, a Self-Organizing Map (SOM) is used to create an over-segmented reference space from the network outputs, where the exposure to different stimuli produces adsorption-desorption paths of activated neurons in real-time. Remarkably, even when a large part of the sensors, 65%, provided anomalous responses, the network offers robust odor identification. Using such a large and easy-to-make platform opens the way to a better understanding and facile implementation of algorithms that trace the architecture of the biological olfaction pathway.

Processing Large Outliers in Arrays of Observations

The interest in large or extreme outliers in arrays of empirical information is caused by the wishes of users (with whom the author worked): specialists in medical and zoo geography, mining, the application of meteorology in fishing tasks, etc. The following motives are important for these specialists: the substantial significance of large emissions, the fear of errors in the study of large emissions by standard and previously used methods, the speed of information processing and the ease of interpretation of the results obtained. To meet these requirements, interval pattern recognition algorithms and the accompanying auxiliary computational procedures have been developed. These algorithms were designed for specific samples provided by the users (short samples, the presence of rare events in them or difficulties in the construction of interpretation scenarios). They have the common property that the original optimization procedures are built for them or well-known optimization procedures are used. This paper presents a series of results on processing observations by allocating large outliers as in a time series in planar and spatial observations. The algorithms presented in this paper differ in speed and sufficient validity in terms of the specially selected indicators. The proposed algorithms were previously tested on specific measurements and were accompanied by meaningful interpretations. According to the author, this paper is more applied than theoretical. However, to work with the proposed material, it is required to use a more diverse mathematical tool kit than the one that is traditionally used in the listed applications.

A Star Pattern Recognition Algorithm Based on the Radial Companion-Circumferential Feature

Attitude measurement is an important core technology of vehicle flight. It is of great significance to ensure the vehicle’s accurate orbit entry and orbit change, high-performance flight, reliable ground communication, high-precision ground observation, and successful completion of various space missions. Star sensor is the core component to realize an autonomous attitude measurement of the vehicle. Autonomous star map recognition is a key technology in star sensor technology, and it is also the focus and difficulty of research. When the star sensor enters the initial attitude acquisition mode, according to the algorithm, the star sensor can quickly obtain the initial attitude and enter the normal working mode. This paper proposes a star pattern recognition algorithm based on the radial companion-circumferential feature with a noise compensation code to address the low recognition rate caused by position noise in the process of constructing star patterns in the traditional star pattern recognition algorithm based on the radial feature. In order to solve the problem of slow matching search speed, a maximum matching number algorithm has been innovatively adopted, which can improve the search efficiency in the process of star pattern recognition. Thus, the capacity of the star pattern recognition feature library is effectively reduced, and the stability and recognition rate of the improved star pattern recognition algorithm are further improved. The improved star pattern recognition algorithm first establishes radial and circumferential feature vectors based on the bit vector, then adds the noise compensation code according to the companion star position error, then modifies the radial and circumferential feature vectors, and finally calculates the minimum similarity difference between the feature vector of the star pattern observed by the star sensor and the feature vector of the navigation star in the feature library to obtain the unique star pattern recognition result. The identification star database adopts the maximum matching number algorithm, which can improve the search efficiency, reduce the amount of redundant matching, and shorten the matching time. The simulation results show that even in the presence of star position and magnitude noise, the improved star pattern recognition algorithm with radial companion-circumferential feature maintains a high recognition rate of more than 97 percent, demonstrating that the algorithm’s robustness is superior to other algorithms. The revised method described in this work outperforms the classic triangle algorithm and the radial feature star pattern recognition algorithm without compensation code in terms of algorithm robustness, recognition success rate, and recognition time.

Delineating unique and discriminant chemical traits in Brazilian and Italian extra-virgin olive oils by quantitative 2D-fingerprinting and pattern recognition algorithms

Brazil is the second-largest importer, after the United States, of extra-virgin olive oils with an internal production that just reaches 0.25 % of the volume imported and consumed. Information about diagnostic chemical signatures characterizing Brazilian oils is lacking, especially as related to geographical markers and key-aroma compounds distribution. This study fills this gap; by comprehensive two-dimensional gas chromatography (GC×GC) combined with headspace solid-phase microextraction (HS-SPME) and parallel detection (MS/FID), diagnostic quali-quantitative signatures of volatiles from Brazilian oils are obtained. Comprehensively mapping of untargeted and targeted (UT) features through fingerprinting algorithms provides effective classification models, benchmarking Brazilian oils (n = 28) vs. a large selection of Italian high-quality extra-virgin olive oils (n = 111). Identitation of Brazilian oils is achieved by combining six different feature selection strategies, the resulting pattern of predictors includes: 1-methoxy-2-propanol, ethyl acetate, (E)-4,8-dimethyl-1,3,7-nonatriene, propanoic acid, limonene, (E)-2-hexenal, hexanoic acid, pentanal, butanoic acid, hexanal, α-pinene, (E)-β-ocimene, and propanal. Compositional differences also are evidenced in composite-class images resembling 2D patterns characteristic of the two production areas. Quantitative fingerprinting, resembling an artificial intelligence smelling machine, characterizes the unique aroma with 1-penten-3-one and (Z)− 3-hexenal, dominating with pungent and green notes, followed by (E)− 2-hexenal (bitter-almond), (E,E)− 2,4-heptadienal (fatty), eucalyptol (herbal), benzaldehyde (almond), hexanal (green), and (E)− 2-hexen-1-ol (green grass).

Network Intelligent Education System Based on the Deep Learning Algorithm

In order to promote the deep integration of education and artificial intelligence and promote the landing of artificial intelligence in the field of education, a method based on deep learning is proposed. Deep learning is a very advanced pattern recognition algorithm, and it can mine potentially valuable knowledge from massive data and provide support for people’s scientific decision-making. Experimental data show that the classification accuracy of deep learning algorithm reaches 99.5%. Even in the case of large data volume and large data categories, the accuracy is still up to 96.3%, which is far higher than the support vector machine algorithm, Bayesian theory algorithm, and K-means algorithm, which can mine and learn knowledge for users more accurately. Conclusion. The network intelligent education system based on deep learning strengthens the automation, intelligence, and interest of learning, so as to ensure the initiative of users in learning.

Simulation of a micron resolution capillary liquid scintillation detector for 14 MeV fusion neutrons

Aiming to improve the spatial resolution of a neutron imaging system (NIS) for 14 MeV fusion neutrons, an ideal micron resolution capillary detector filled with a high optical index liquid scintillator was simulated. A threshold for each capillary pixel and a threshold for each cluster were applied to suppress the gamma-induced background. In addition, by using a pattern recognition algorithm and an optimized Hough transform, the accuracy of determining the neutron impinging positions and the dynamic range of this detector were enhanced. For an ideal capillary array detector, the spatial resolution is expected as one capillary size of 20μm. The dynamic range of ∼1000 is reachable while the accuracy of neutron impinging position determination keeps better than 85%. The ionization quenching, light sharing and energy resolution of the detector were applied to the simulated data to understand the capillary array detector.

Computing graph edit distance on quantum devices

Distance measures provide the foundation for many popular algorithms in Machine Learning and Pattern Recognition. Different notions of distance can be used depending on the types of the data the algorithm is working on. For graph-shaped data, an important notion is the Graph Edit Distance (GED) that measures the degree of (dis)similarity between two graphs in terms of the operations needed to make them identical. As the complexity of computing GED is the same as NP-hard problems, it is reasonable to consider approximate solutions. In this paper, we present a QUBO formulation of the GED problem. This allows us to implement two different approaches, namely quantum annealing and variational quantum algorithms, that run on the two types of quantum hardware currently available: quantum annealer and gate-based quantum computer, respectively. Considering the current state of noisy intermediate-scale quantum computers, we base our study on proof-of-principle tests of their performance.

Determination of oil pollutants by three-dimensional fluorescence spectroscopy combined with improved pattern recognition algorithm

Petroleum refineries are one of the main sources of hazardous air pollutants, so the accurate determination of petroleum pollutants is of great significance to maintain ecological balance. In this study, three-dimensional (3D) fluorescence spectroscopy combined with pattern recognition algorithm is adopted to distinguish the composition and content of oil pollutants efficiently and accurately. Three hundred samples of kerosene, diesel, and gasoline mixed solutions with different concentrations are prepared. The principal component analysis is used to extract the optimal feature variables, and the correlation coefficient method is used to obtain eight groups of principal component features in the spectra. The dimension is selected as 8, and the principal component score is calculated, which is used as the input data of the extension neural network. Next, the pattern recognition method is improved, and the designed neural network has functions of both resolution and measurement. The results of neural network pattern recognition are used as the input of the concentration network. The first 270 samples are used as the training samples to train the network model, and the remaining 30 samples are used as test samples, which are applied to the input layer of the trained neural network. The relative fluorescence intensity, relative slope, and comprehensive background parameters are used as the input parameters, and the extension neural network is used for pattern recognition and evaluation of oil pollutants. The experimental results show that the average recognition rate of the improved pattern recognition algorithm for oil pollutants is 98.43%, and the average recovery rate of concentration is 98.67%. Further, the average time for pattern recognition is 1.53 s, while the parallel factor analysis algorithm takes 2.89 s. This suggests that the improved extension neural network is an effective and reliable pattern recognition method for the identification of mixed oil pollutants.

Verbal interaction pattern analysis in clinical psychology.

Recent developments in pattern analysis research have made this methodology suitable for the study of the processes that are set in motion in psychological interventions. Outcome research, based on the comparison between clinical results from treatment and control groups, has leveraged our empirical knowledge about the efficacy of psychological interventions. However, these methods of research are not precise enough for the analysis of these processes. On the contrary, pattern analysis could be a powerful tool to study moment-to-moment interactions typical of psychological interventions. This is methodology is relevant because clinical psychology is experiencing a paradigm shift from a protocol for syndrome perspective to a principle-based and person-centered intervention. This evidence-based, theory-grounded, and process-oriented paradigm of clinical intervention needs new research methods to thrive (i.e., pattern analysis). The analysis of the therapeutic relationship built into the verbal interaction between the clinician and the client is one of the cornerstones of this new era of research. So, the purpose of this article is three-fold: (1) to discuss the role of the verbal interaction pattern analysis in the clinical context to the development of the principle-based clinical psychology, (2) to analyze the patterns of verbal interaction in a clinical case, and (3) to compare the results using two different methods. To reach these purposes, using the observational methodology, we have coded the verbal interaction of 16 clinical sessions with a person diagnosed with a borderline personality disorder. We have analyzed the data using sequential analysis (GSEQ) and pattern recognition algorithms (i.e., T-Pattern detection). We have been able to detect typical patterns during different phases of psychological intervention (i.e., evaluation, explanation, treatment, and consolidation). Finally, the conceptual, methodological, and empirical implications of this study will be discussed within the realms of pattern analysis research and principle-based clinical psychology.

Headspace GC/MS and fast GC e-nose combined with chemometric analysis to identify the varieties and geographical origins of ginger (Zingiber officinale Roscoe)

Food authenticity regarding different varieties and geographical origins is increasingly becoming a concern for consumers. In this study, headspace gas chromatography-mass spectrometry (HS-GC–MS) and fast gas chromatography electronic nose (fast GC e-nose) were used to successfully distinguish the varieties and geographical origins of dried gingers from seven major production areas in China. By chemometric analysis, a distinct separation between the two varieties of ginger was achieved based on HS-GC–MS. Furthermore, flavor information extracted by fast GC e-nose realized the discrimination of geographical origins, and some potential flavor components were selected as important factors for origin certification. Moreover, several pattern recognition algorithms were compared in varietal and regional identification, and random forest (RF) led to the highest accuracies for discrimination. Overall, a rapid and precise method combining multivariate chemometrics and algorithms was developed to determine varieties and geographical origins of ginger, and it could also be applied to other agricultural products.

Acoustic emission-based monitoring of fatigue damage in CFRP-CFRP adhesively bonded joints

Adhesively bonded joints are being increasingly applied in modern structures. However, manufacturing defects and particularly harsh operative conditions may cause local debonding and catastrophic failures. Structural health monitoring (SHM) and non-destructive testing (NDT) procedures are then needed to evaluate the in-service structural integrity of adhesively bonded joints. In this research, an adhesively bonded single-lap joint, both adherends of which are manufactured using a carbon fibre-reinforced polymer (CFRP) composite, is subjected to constant amplitude fatigue loading. During such a test, the integrity and damage condition of the joint is continuously monitored using acoustic emission (AE), while the test itself is periodically interrupted in order to apply micro-computed tomography (μCT) to the specimen, with the aim of investigating the real features of the developing fatigue damage. The results show that, after suitable elaboration and filtering by means of pattern recognition algorithms, acoustic emission-based monitoring allows for effective identification and characterisation of the development of fatigue damage in adhesively bonded joints.

Automatic Tracking of Weak Acoustic Targets within Jamming Environment by Using Image Processing Methods

Bear time records, which are the accumulations of spatial spectrum estimates on the time axis, are often employed for passive sonar information processing. Multi-target jamming is a common difficulty in this approach due to the constraints of Rayleigh limit, and neither the conventional beamforming (CBF) nor minimum variance distortionless response (MVDR) technique can handle it well. This work presents a post-processing tracking framework based on visual pattern recognition algorithms to track weak acoustic targets within jamming environments, which includes target motion analysis, matched filtering, and principal component analysis-based denoising, and we call this ‘P-Gabor’ algorithm. The simulations and sea-trial experiments show that the proposed method can track a weak target successfully under −23 dB (signal-to-interference ratio) SIR, which is more effective than the references, especially in terms of using real-world data from sea trials. We further demonstrate that the method also has stable tracking performance at even −25 dB SNR (signal-to-noise ratio) circumstances.