Feature Point Recognition Research Articles

Body landmarks and genetic algorithm-based approach for non-contact detection of head forward posture among Chinese adolescents: revitalizing machine learning in medicine

Addressing the current complexities, costs, and adherence issues in the detection of forward head posture (FHP), our study conducted an exhaustive epidemiologic investigation, incorporating a comprehensive posture screening process for each participant in China. This research introduces an avant-garde, machine learning-based non-contact method for the accurate discernment of FHP. Our approach elevates detection accuracy by leveraging body landmarks identified from human images, followed by the application of a genetic algorithm for precise feature identification and posture estimation. Observational data corroborates the superior efficacy of the Extra Tree Classifier technique in FHP detection, attaining an accuracy of 82.4%, a specificity of 85.5%, and a positive predictive value of 90.2%. Our model affords a rapid, effective solution for FHP identification, spotlighting the transformative potential of the convergence of feature point recognition and genetic algorithms in non-contact posture detection. The expansive potential and paramount importance of these applications in this niche field are therefore underscored.

DLATA: Deep Learning-Assisted transformation alignment of 2D brain slice histology

Accurate alignment of brain slices is crucial for the classification of neuron populations by brain region, and for quantitative analysis in in vitro brain studies. Current semi-automated alignment workflows require labor intensive labeling of feature points on each slice image, which is time-consuming. To speed up the process in large-scale studies, we propose a method called Deep Learning-Assisted Transformation Alignment (DLATA), which uses deep learning to automatically identify feature points in images after training on a few labeled samples. DLATA only requires approximately 10% of the sample size of other semi-automated alignment workflows. Following feature point recognition, local weighted mean method is used as a geometrical transformation to align slice images for registration, achieving better results with about 4 fewer pixels of error than other semi-automated alignment workflows. DLATA can be retrained and successfully applied to the alignment of other biological tissue slices with different stains, including the typically challenging fluorescent stains. Reference codes and trained models for Nissl-stained coronal brain slices of mice can be found at https://github.com/ALIGNMENT2023/DLATA.

Analysis of feature point matching technology in SLAM based on binocular vision

With rapidly developing computing technology, intelligent robots have been used extensively in many areas of life, improving work efficiency and lowering labor costs. At the same time, with the increasing number of robot application scenarios, the demand for robot intelligence has also increased. Computer vision-based simultaneous localization and mapping (SLAM) technology is a key technology that helps robots achieve real-time positioning and navigation in order to improve their intelligence. Feature point matching is an important component of this technology. This paper mainly analyzes the current development status of feature point matching technology in visual SLAM. Firstly, a brief introduction was given to the three widely used SLAMs, and the advantages and disadvantages of different SLAM technologies were analyzed. Second, a brief presentation and explanation of the working principle of binocular SLAM technology were provided. Then, the basic principles and key points of feature point recognition and matching techniques in several different algorithms for visual SLAM were summarized. Analyzed the advantages and disadvantages of various algorithms, as well as improvements based on this algorithm. Finally, summarize and provide suggestions for the future development of visual SLAM technology.

Monocular vision measurement technology applied in dynamic compaction ramming settlement monitoring

Ramming settlement is a key index to reflect the quality of dynamic compaction. It mainly relies on manual monitoring, which seriously affects construction efficiency, and safety cannot be guaranteed. In this paper, an algorithm Perspective 3-Point Rammer Pose Estimation (P3P-RPE) based on monocular vision measurement is proposed, which can be carried out simultaneously with the construction, effectively avoiding the problem that the dynamic compaction operation needs to be terminated during manual monitoring. Then the error propagation model of P3P-RPE is proposed, and the influencing factors are analyzed comprehensively. The results show that the accuracy can reach 30 mm, which is less than the upper limit of 50 mm specified in the specification. And pitch angle of camera θ and feature point recognition have a great influence on the results. This method has good practicability in ramming settlement monitoring, and can provide reference for the monitoring of other similar projects.

Robust and Fast Normal Mollification via Consistent Neighborhood Reconstruction for Unorganized Point Clouds.

This paper introduces a robust normal estimation method for point cloud data that can handle both smooth and sharp features. Our method is based on the inclusion of neighborhood recognition into the normal mollification process in the neighborhood of the current point: First, the point cloud surfaces are assigned normals via a normal estimator of robust location (NERL), which guarantees the reliability of the smooth region normals, and then a robust feature point recognition method is proposed to identify points around sharp features accurately. Furthermore, Gaussian maps and clustering are adopted for feature points to seek a rough isotropic neighborhood for the first-stage normal mollification. In order to further deal with non-uniform sampling or various complex scenes efficiently, the second-stage normal mollification based on residual is proposed. The proposed method was experimentally validated on synthetic and real-world datasets and compared to state-of-the-art methods.

Non-contact clothing anthropometry based on two-dimensional image contour detection and feature point recognition

Developing the technology of estimating human body size from two-dimensional images is the key to realising more digitalization and artificial intelligence in the textile and garment industry. Therefore, this paper is an in-depth study of estimating body sizes from two-dimensional images in a self-collected database of human body samples. First, the artificial thresholds in the Canny edge operator were replaced by adaptive thresholds. The improved Canny edge operator was combined with mathematical morphology so that it could detect a clear and complete single human contour. Then a joint point detection algorithm based on a convolution neural network and human proportion is proposed. It can detect human feature points with different body proportions. Finally, front and side images and manual body measurements of 122 males aged 18–22 years were collected as the human sample database, calculating the length and fit of the girth size. Compared with manual body measurement data, the error of human length and girth size parameters within the national standard range of –1.5 ~ 1.5 cm can reach 91% on average. This study provides an accurate and convenient anthropometric method for digital garment engineering, which can be used for online shopping and garment customization, and has a certain practical value.

Multi-Morphological Pulse Signal Feature Point Recognition Based on One-Dimensional Deep Convolutional Neural Network

Radial pulse signals are produced by the periodic ejection of blood from the heart, and physiological and pathological information of the human body can be analyzed by extracting the time-domain characteristics of pulse waves. However, since pulse signals are weak physiological signals on the body surface and complex, the acquisition of pulse characteristics using the traditional curvature method will produce a large error, which cannot meet the needs of pulse wave analysis in current clinical practice. To solve this problem, a multi-morphological pulse signal feature recognition algorithm based on the one-dimensional deep convolutional neural network (1D-DCNN) model is proposed. We used the multi-channel pulse diagnosis instrument independently developed by the team to collect radial pulse signals under continuous pressure of the test subjects and collected 115 subjects and extracted a total of 1300 single-cycle pulse signals and then divided these pulse signals into 6 different forms. Five types of pulse signal time-domain feature points were labeled, and five independent feature point datasets were labeled and formed five customized neural network models that were generated to train and identify the pulse feature point datasets independently. The results show that the correction coefficient (Radjusted2) of the multi-class pulse signal processing algorithm proposed in this paper for each type of feature point recognition reaches more than 0.92. The performance is significantly better than that of the traditional curvature method, which shows the accuracy and superiority of the proposed method. Therefore, the multi-class pulse signal characteristic parameter recognition model based on the 1D-DCNN model proposed in this paper can efficiently and accurately identify pulse time-domain characteristic parameters, which can be applied to discriminate time-domain pulse information in clinical practice and assist doctors in diagnosis.

Real-Time Robotic Mirrored Behavior of Facial Expressions and Head Motions Based on Lightweight Networks

The ability of a humanoid robot to imitate facial expressions with simultaneous head motions is crucial to natural human–robot interaction. This mirrored behavior from human beings to humanoid robots has high demands of similarity and real-time performance. To fulfill these needs, this article proposes a real-time robotic mirrored behavior of facial expressions and head motions based on lightweight networks. First, a humanoid robot that can change the state of its facial organs and neck through servo displacement is developed to achieve the mirrored behavior of facial expressions and head motions. Second, to overcome the high latency caused by deep learning models running in embedded devices, a lightweight deep learning network is constructed for detecting facial feature points, which can reduce model size and improve running speed without affecting the performance of the model. Finally, a mapping relationship of 68 facial feature points to optimal servo displacements is established to realize the mirrored behavior from human beings to humanoid robots. The experimental results show that the facial feature point recognition method based on the lightweight model performs better than other state-of-the-art methods, and our head motion tracking method can maintain high accuracy compared with the gold standard optical motion capture system NOKOV. Overall, our method ensures the accurate and real-time generation of robot mirrored behavior and has a certain reference value for the efficient and natural interaction between humans and robots.

Research on vehicle speed identification method based on time interpolation method and feature point recognition

Based on camera calibration in detection area, an improved method of time interpolation for vehicle speed identification is proposed. Firstly, calibration of camera parameters during photogrammetry is carried out through preset rectangular detection area. Secondly, the key frames are extracted from the video images and the target vehicle feature points and their image coordinates are identified by the image morphological processing. Then the actual coordinates of the characteristic points are determined through the coordinate conversion. Finally, the improved time interpolation method considering the acceleration between frames is used to identify the vehicle speed, so as to improve the accuracy of the vehicle speed measurement. The results proved that the vehicle speed calculated by improved time interpolation method is more reliable, and the error is reduced by 0.7%.

Design of pulse and respiration monitoring system based on fiber optic sensing and VMD-FPR processing algorithm

We propose a pulse and respiration monitoring system based on fiber optic sensing, to achieve accurate monitoring of human pulse and respiration rate with wearable and high comfort. We designed a flexible packaging structure -- to embed the fiber Bragg grating sensors (FBGs) fiber optic sensor into a nylon bandage, which constitutes a wearable real-time monitoring system that can acquire pulse and respiration rates simultaneously. The system uses the Variational Mode Decomposition - Feature Point Recognition (VMD-FPR) signal processing algorithm to demodulate and analyze the FBG echo signal to invert the pulse and respiration rate. The experiments were conducted on human heart rate and respiration rate in different states. The experimental results showed that the error range of respiration rate was within ±0.2 bpm, and the accuracy of the respiration experiment feature point recognition algorithm was 93.29 %. The error range of heart rate was within ±2 bpm, and the accuracy of the pulse experiment main wave feature point recognition algorithm was 99.28 %. Thus, the system can accurately collect and analyze the feature points in the respiratory and pulse wave signals. The information obtained from the experiments is used to build an information base, which provides data support for establishing the analysis model of pulse characteristics and respiratory characteristics of the detectors, and to achieve the function of real-time monitoring of human physiological parameters.

Design of a Real-Time Movement Decomposition-Based Rodent Tracker and Behavioral Analyzer Based on FPGA

In this work, a real-time movement decomposition-based automatic behavior analysis system is proposed. The proposed method is evaluated on a Morris water maze experiment with rat. The system integrated a high-speed tracker, a feature extraction module, and a classifier on a Xilinx Ultra-96 single computer board with a field-programmable gate array (FPGA). The high-speed tracker includes a movement predictor based on motion features and pose features, and a fast-checking algorithm to evaluate whether the tracking result is correct or not. A faster region-based convolutional neural network (Faster-RCNN)-based detector is used for initializing the tracker. The feature extraction module integrated a CNN-based feature point recognition network and a motion feature computing module. Motion and pose feature are calculated with the result of the feature points. The proposed system achieves a 66.2-frames/s processing speed, and the average point detection error is 4.01 pixels. For each video, the classification and regression tree (CART)-based classifier gives an explorative stage in five categories, which represents the rat’s strategy of exploration in Morris water maze task. This work focuses on a real-time local computing system design on small-scale datasets.

Feature Points Recognition of Computerized Numerical Control Machining Tool Path Based on Deep Learning

In the processes of feed rate planning, interpolation and tool path optimization of Computerized Numerical Control (CNC) machining, feature point recognition of tool path is an essential operation. Accurate identification of feature points is the premise for partitioning the tool path that affects the accuracy and efficiency of CNC machining. Current research on feature point recognition of tool path mainly relies on the hand-crafted approach, of which the result is sensitive to the values of some manually defined threshold. Therefore the approach heavily relies on the human experience. In this paper, a novel deep learning-based approach is presented that can automatically and precisely recognize the feature point of the tool path. A set of geometric descriptors is first defined for each Cutter Location (CL) point and then conversed to images that feed to the deep learning pipeline. A residual learning-based Convolutional Neural Network (CNN) called Feature Point CNN (FP-CNN) is designed that takes the conversed images as input and the recognized results as output. Extensive experiments on some industrial parts are conducted to validate the effectiveness and the advantage of the proposed network. Results show that the proposed approach has good performance in identifying accuracy and recall, which is much superior to two types of benchmarks and does not involve any human intervention.

Research on obstacle avoidance gait planning of quadruped crawling robot based on slope terrain recognition

PurposeAiming at the problem that quadruped crawling robot is easy to collide and overturn when facing obstacles and bulges in the process of complex slope movement, this paper aims to propose an obstacle avoidance gait planning of quadruped crawling robot based on slope terrain recognition.Design/methodology/approachFirst, considering the problem of low uniformity of feature points in terrain recognition images under complex slopes, which leads to too long feature point extraction time, an improved ORB (Oriented FAST and Rotated BRIEF) feature point extraction method is proposed; second, when the robot avoids obstacles or climbs over bumps, aiming at the problem that the robustness of a single step cannot satisfy the above two motions at the same time, the crawling gait is planned according to the complex slope terrain, and a robot obstacle avoidance gait planning based on the artificial potential field method is proposed. Finally, the slope walking experiment is carried out in the Robot Operating System.FindingsThe proposed method provides a solution for the efficient walking of robot under slope. The experimental results show that the extraction time of the improved ORB extraction algorithm is 12.61% less than the original ORB extraction algorithm. The vibration amplitude of the robot’s centroid motion curve is significantly reduced, and the contact force is reduced by 7.76%. The time it takes for the foot contact force to stabilize has been shortened by 0.25 s. This fact is verified by simulation and test.Originality/valueThe method proposed in this paper uses the improved feature point recognition algorithm and obstacle avoidance gait planning to realize the efficient walking of quadruped crawling robot on the slope. The walking stability of quadruped crawling robot is tested by prototype.

Inspection of the Underwater Industrial Objects Using AUV Based on the Method of Feature Points Video Recognition

Automated inspection of seabed industrial equipment with the help of AUV assumes its navigational referencing to the inspection object with sub-met er accuracy. That accuracy cannot be achieved on the base of standard onboard navigation equipment. It is proposed to use the method of video recognition of feature points (FP) on the inspected equipment to solve this problem. The FPs are preliminarily determined on the models of objects and downloaded into the AUV control system in the form of three-dimensional sets (clouds). Ensuring stable high accuracy during continuous movement of AUV in inspected objects space is realized through regular updating of the coordinate referencing to the objects. The inspection procedure using this method consists of the following steps:- the AUV mission specification (including downloading the FPs descriptions),- launching the AUV and preliminary approach to the inspected objects,- determining the optimal sequence for bypassing the inspected objects,- performing the searching procedure for the next FP and navigational reference to it,- subsequent inspection of the specified fragments of equipment.At the same time, the logic of the mission execution provides for a dynamic change in the inspection plan in case of failure when any of the planned FP are detected. The paper presents the AUV control system algorithms that implement the described inspection scenario. The approach correctness was investigated using the mathematical modeling methods.

Research on lithium battery sorting method based on image adaptive recognition of feature points

Research on lithium battery sorting method based on image adaptive recognition of feature points

Research on lithium battery sorting method based on image adaptive recognition of feature points

When sorting batteries in factories, fixed feature points in a discharge voltage platform are usually used as a sorting basis with expert experience, causing poor accuracy in battery sorting. This study proposes a method based on feature point image adaptive recognition (FPIAR) and combined a semi-supervised fuzzy C-means (SFCM) clustering algorithm to achieve battery sorting. Accordingly, FPIAR obtained the multi-dimensional feature coordinate dataset by adaptively identifying the discharge voltage curve of each battery. The centre point of the accumulation of coordinates in the dataset was found by mean-shift algorithm and was used as the feature point of battery sorting voltage. In addition, the SFCM clustering algorithm was used to process the feature point of the sorting voltage to sort the batteries into groups. The simulation results and tests indicated that the SFCM clustering algorithm based on FPIAR processing can improve the accuracy of battery sorting.

Facial feature point recognition method for human motion image using GNN

<abstract> <p>To address the problems of facial feature point recognition clarity and recognition efficiency in different human motion conditions, a facial feature point recognition method using Genetic Neural Network (GNN) algorithm was proposed. As the technical platform, weoll be using the Hikey960 development board. The optimized BP neural network algorithm is used to collect and classify human motion facial images, and the genetic algorithm is introduced into neural network algorithm to train human motion facial images. Combined with the improved GNN algorithm, the facial feature points are detected by the dynamic transplantation of facial feature points, and the detected facial feature points are transferred to the face alignment algorithm to realize facial feature point recognition. The results show that the efficiency and accuracy of facial feature point recognition in different human motion images are higher than 85% and the performance of anti-noise is good, the average recall rate is about 90% and the time-consuming is short. It shows that the proposed method has a certain reference value in the field of human motion image recognition.</p> </abstract>

A novel angle extremum maximum method for recognition of pulse wave feature points

Background and objectivesPulse wave is one of the biomedical signals that has been studied over the past years. Accurate recognition of feature points is the basis of verifying the connections between pulse waves and certain diseases. Therefore, the aim of the study is to discuss the use of angle mapping on feature points recognition. MethodsThe mathematical method is based on the application of angle curve with parameter " k " on pulse wave. The data used is collected by PVDF sensor. Approximate curve and mathematical model are used for the discussion of the influence of parameter k and pulse wave amplitude by numerical calculation. The conclusion drawn from the numerical solution is that when k changes to maximize the angle extremum value, the corresponding position of angle extremum point is the feature point position. For the sampling rate f = 455Hz in this paper, k can be taken from 5 to 15. ResultsWe present the recognition results of unobvious feature points based on the "angle extremum maximum method" and corresponding angle values. The results are compared with traditional methods and the determination of angle threshold value is discussed. ConclusionsThis method can be used for accurate and efficient feature points identification, and it can be better applied to pulse waves with noise or unobvious feature points.

Design of Chinese Character Coded Targets for Feature Point Recognition Under Motion-Blur Effect

Coded targets have been widely used as a type of active visual feature points in fields such as close-range photogrammetry, robot navigation, 3D reconstruction, and augmented reality. However, coded targets in degraded images, such as images with motion blur effect, are hard to recognize. To this end, a set of novel Chinese character coded targets (CCTs) is designed and tested. A CCT is a square visual marker, shown as a relative small circular feature overlaid in the middle of a square Chinese character. A white circular ring is embedded in a black circle concentrically to serve as the circular feature, which facilitates extraction of the center point of the marker for localization. Whereas the distinctive peripheral Chinese character appearance of each CCT is utilized for identification. By synthesizing simulated CCTs with different degrees of motion blur and various postures with the real background images, a Faster Region-based Convolutional Neural Network (Faster R-CNN) is trained to locate and recognize the CCTs in motion blurred images. Experimental results on both artificial and actual motion blurred images demonstrate the superiorities of the designed CCTs as well as the proposed localization and recognition pipeline.

Design of fast digital spectrum stabilization method for UAV radiometry system

The drone radiation measurement system has a sharp increase in demand for environmental radiation monitoring due to its easy manoeuvrability, low aerial survey cost, high patrol efficiency and no personnel safety concerns. In order to adapt to the fast, flexible and efficient characteristics of the drone, the detector for the Unmanned Aerial Vehicle (UAV) radiation measurement generally consists of a plurality of small-sized detectors to form a detection array. Due to the low single crystal count rate and the influence of electronic system temperature and packaging, many detectors of the UAV radiation measurement system will have “spectral drifts”, which affects the quality of the synthesized spectrum and affects the accuracy of the whole system. Aiming at the characteristics of UAV spectrum measurement system, a three feature point recognition model is established, and the K-40 feature peak (1.46MeV) is approximated by reverse piecewise summation. Thus, the backscattering peaks and characteristic peaks K-40 of the measured spectrum can be obtained to achieve fast spectrum stabilization. Finally, through Gaussian Mixture Model (GMM) peak estimation and experimental validation, the method can obtain the characteristic peak position accurately, and the time between instrument start-up and automatic spectrum stabilization is less than 2 minutes.