Distance Of Point Research Articles

RFCS-YOLO: Target Detection Algorithm in Adverse Weather Conditions via Receptive Field Enhancement and Cross-Scale Fusion

The paper proposes a model based on receptive field enhancement and cross-scale fusion (RFCS-YOLO). It addresses challenges like complex backgrounds and problems of missing and mis-detecting traffic targets in bad weather. First, an efficient feature extraction module (EFEM) is created. It reconfigures the backbone network. This helps to make the receptive field better and improves its ability to extract features of targets at different scales. Next, a cross-scale fusion module (CSF) is introduced. It uses the receptive field coordinate attention mechanism (RFCA) to fuse information from different scales well. It also filters out noise and background information that might interfere. Also, a new Focaler-Minimum Point Distance Intersection over Union (F-MPDIoU) loss function is proposed. It makes the model converge faster and deals with issues of leakage and false detection. Experiments were conducted on the expanded Vehicle Detection in Adverse Weather Nature dataset (DWAN). The results show significant improvements compared to the conventional You Only Look Once v7 (YOLOv7) model. The mean Average Precision (mAP@0.5), precision, and recall are enhanced by 4.2%, 8.3%, and 1.4%, respectively. The mean Average Precision is 86.5%. The frame rate is 68 frames per second (FPS), which meets the requirements for real-time detection. A generalization experiment was conducted using the autonomous driving dataset SODA10M. The mAP@0.5 achieved 56.7%, which is a 3.6% improvement over the original model. This result demonstrates the good generalization ability of the proposed method.

Volume measurement and calculation of sessile droplets on inclined planes and spherical external surfaces

This paper puts forward a novel measurement and calculation method for sessile droplets' volumes on inclined planes and spherical external surfaces. The volume of a droplet on an inclined plane or on a spherical external surface, which has a fictitious inclined plane, is determined by its shape factor (β), the maximum distance (d) of the contour points away from the inclined plane, and the plane inclination angle θinc. Using the circular arc differential method of the Bashforth–Adams equation and the image processing technique, the d and β are obtained or calculated out. The theoretical model and its algorithm are set up to calculate the droplet volume. In addition, we do experiments to validate the accuracy of the theoretical model. The experimental results show good agreement with those calculated by the theoretical model. This work is significant and can be used to measure and control volumes of droplets in microfluidic and biomedical technologies.

Fully guided system for position-predictable autotransplantation of teeth: A randomized clinical trial.

Autotransplantation of teeth (ATT) is a viable biological method for addressing dental defects. The objective was to achieve occlusal reconstruction-orientated ATT to enhance functionality and obtain optimal location and adjacency. This study proposes a new concept of a guide (a fully guided system) to achieve position-predictable ATT. This study included 30 transplants in 29 patients who were randomly divided into two groups during the initial examination using a single-blind method. The experimental group comprised 15 transplants in 14 patients who underwent surgery using a fully guided system. Fifteen transplants were performed on 15 patients in the control group using only a replica. The mean number of repeated teeth autotransplantation attempts and mean preparation time exhibited no significant differences. The mean follow-up period was 1 year. Twenty-six transplants were followed up clinically, while four were only followed up by phone calls. The experimental group demonstrated a 100% success rate, whereas the control group exhibited a 70% success rate, with an overall 100% survival rate. 3D analysis: Deviation of both angle and distance was calculated for evaluation of accuracy. The mean centre deviation of teeth in experiment group was 0.79 mm (95% confidence interval [CI], 0.51-1.08) compared to 1.79 mm (95% CI, 1.10-2.47) in control group (p < .05). The mean of average index points distance deviation showed a significant difference. The mean angle deviation of long axis in experiment group was 3.87° (95% CI, 2.25-5.49) compared to 12.98° (95% CI, 8.64-17.31) in control group (p < .05). The angle deviation of long axis in mesio-distal direction, buccal-lingual direction and rotation showed a significant difference also. The results of the analysis indicated that the fully guided system increased the predictability of achieving an ideal transplant position with intuitive data, establishing a solid foundation for optimal dental functional recovery and restoration. However, this study had limitations regarding the generalisability of the design of the optimal location for transplantation and the equipment requirements for clinical procedures. Therefore, further research on the screening of indications for guided systems and location planning is necessary. This study was registered on the Chinese Clinical Trial Registry (Registration Number: ChiCTR 2300074646).

Data-driven optimisation of process parameters for reducing developed surface area ratio in laser powder bed fusion

Abstract Optimising process parameters is a key category of approaches to improve the surface quality of laser powder bed fusion parts. So far, many optimisation methods have been presented, which provide effective ideas and approaches for improving surface quality. However, these methods all focus on the improvement of $$R_a$$ R a , $$S_a$$ S a , $$R_{sk}$$ R sk , or $$R_{\Delta q}$$ R Δ q . These parameters are sufficient for most applications. They are however not suitable for applications where functional performance is linked with surface area. To quantify the quality of surfaces for these applications, the developed surface area ratio ( $$S_{dr}$$ S dr ) is a more appropriate parameter as it can be used to quantitatively express the exposed functional surface area. In this paper, a data-driven method for optimising five process parameters, namely layer thickness, laser power, hatch spacing, point distance, and exposure time, to reduce the developed surface area ratio of laser powder bed fusion parts is proposed. Firstly, experiments are designed, and actual build and measurement experiments are conducted to acquire a fixed amount of data. A Bayesian ridge regression model for predicting a developed surface area ratio from the five process parameters is then trained and tested and compared with several other machine learning models using the acquired data. After that, optimisation of the five process parameters to reduce developed surface area ratio is carried out using the genetic algorithm, in which the objective function values (developed surface area ratios) are predicted using the established Bayesian ridge regression model. Finally, an additional actual build and measurement experiment is conducted to validate the optimisation. The testing results show that the Bayesian ridge regression model can obtain an average R $$^2$$ 2 score of 0.77 and an average mean absolute error of 8.48 on the testing dataset. The validation results suggest that the developed surface area ratios generated by the optimisation are relatively small, and on average, they are 52.11% smaller than the developed surface area ratio under the process parameters recommended by the used laser powder bed fusion system.

CMM Influence Factors and Uncertainty Associated with Length Measurement

This paper is concerned with coordinate measuring machine (CMM) uncertainty evaluation, in particular, the uncertainties associated with point clouds and distances derived from the point cloud. The uncertainty evaluation approach is model-based following the principles of the Guide to the Expression of Uncertainty in Measurement and the law of the propagation of uncertainty. The paper considers a range of CMM influence factors and derives an explicit dependence for the point cloud data coordinates on the influence factors, allowing uncertainties associated with the influence factors to be propagated through to point cloud uncertainties. The paper describes the use of Gaussian processes to model kinematic and probing errors using a small number of statistical hyper-parameters. These models permit an explicit statement of the uncertainty associated with point clouds and length measurement, enabling the latter to be compared directly with a statement of the maximum permissible error in length measurement. The uncertainty evaluation methodology is direct in that it requires no optimisation nor Monte Carlo simulations.

Biomechanical evolutionary advantage of flexor pollicis longus accessory head and topographical association with neurovascular structures.

The human upper limb has undergone various evolutionary myologic changes, accompanied by corresponding modifications in the anatomical course of neurovascular structures. In this study, we aimed to elucidate the emergence of the accessory head of the flexor pollicis longus (AHFPL) muscle as a beneficial biomechanical evolutionary development and its topographical relationship with adjacent neurovascular structures. In this pursuit to understand this phenomenon, dissections were conducted on sixty-two upper limbs from thirty-one cadavers. We found a strong positive correlation between the length of the AHFPL and the distance of the branching point of the brachial artery, ulnar artery, and median nerve into the anterior interosseous nerve. Additionally, we noted a significant negative correlation between the width of the AHFPL muscle and the angle formed between the long axes of the FPL and AHFPL. This increases the angle between the resultant vector and the FPL muscle, decreasing the angle between the AHFPL and the resultant force. Consequently, the resulting force shifts from the initial radial position to a slightly ulnar side, balancing the muscle forces at the same point. The authors hypothesize that this resultant vector provides a biomechanical advantage for the thumb, enhancing emphatic grip precision, providing extra power, and enabling meticulous cupping grip when using tools. Rather than considering the presence of AHFPL as a mere anatomical variation, it should be viewed as an advantageous evolutionary biomechanical development.

A Laser-Based SLAM Algorithm of the Unmanned Surface Vehicle for Accurate Localization and Mapping in an Inland Waterway Scenario

It is important to improve the localization accuracy of the unmanned surface vehicle (USV) for ensuring safe navigation in an inland waterway scenario. However, the localization accuracy of the USV is affected by the limited availability of global navigation satellite system signals, the sparsity of feature points, and the high scene similarity in inland waterway scenarios. Therefore, this paper proposes a laser-based simultaneous localization and mapping (SLAM) algorithm for accurate localization and mapping in inland waterway scenarios. Inertial measurement unit (IMU) data are integrated with lidar data to address motion distortion caused by the frequent motion of the USV. Subsequently, a generalized iterative closest point (GICP) algorithm incorporating rejection sampling is integrated to enhance the accuracy of point cloud matching, involving a two-phase filtering process to select key feature points for matching. Additionally, a mixed global descriptor is constructed by combining point cloud intensity and distance information to improve the accuracy of loop closure detection. Experiments are conducted on the USV-Inland datasets to evaluate the performance of the proposed algorithm. The experimental results show that the proposed algorithm generates accurate mapping and significantly improves localization accuracy by 25.6%, 18.5%, and 23.6% compared to A-LOAM, LeGO-LOAM, and ISC-LOAM, respectively. These results demonstrate that the proposed algorithm achieves accurate localization and mapping in an inland waterway scenario.

ICGM-YOLO: UAV remote sensing small target detection algorithm based on feature fusion enhancement module and ghost convolution

Abstract Remote sensing images captured at high altitudes are affected by background noise and low resolution, leading to suboptimal performance of commonly used small object detection algorithms. This paper introduces ICGM-YOLO, an enhanced approach designed to improve the precision of detecting small objects within remote sensing images. The methodology refines YOLOv8 for optimal performance in this context. First, within the feature extraction network, iterative attention feature fusion combines feature maps and integrates features from different layers. Second, the convolutional layer and C2f module are replaced with the GhostConv module and C3Ghost to extract features. Finally, this paper replaces CIoU with the MFIoU loss function, which combines the new bounding box MPDIoU based on minimum point distance with Focaler-IoU. This replacement accelerates model convergence and enhances detection recall. Experimental results from the Visdrone2019 dataset indicate that ICGM-YOLO outperforms the original YOLOv8 by achieving an 8.1% improvement in mAP0.5 detection rate and a 5.3% increase in mAP0.5:0.95 detection rate. Additionally, ICGM-YOLO reduces the parameter count by 48.2% and decreases computational complexity by 45.2%.

Multilayer CVD Graphene Coatings Developed with Suitable Geometrical Parameters for Improved Corrosion Resistance of Ni and a Ni-Cu Alloy in Chloride Environment.

Graphene layers on metals developed by chemical vapor deposition (CVD) possess characteristics of an ideal coating for durable corrosion resistance of industrial infrastructure, by a green approach. However, developing coatings of consistent quality is a challenge and its circumvention requires manipulation of CVD process parameters. The great variability (from remarkable to little) in graphene's ability as a corrosion barrier coating is attributed to the extent of defects/non-uniformity of graphene. Minimizing the defects by controlling a few CVD process parameters (chemistry, pressure, temperature, and flowrate of the precursor, and post-CVD cooling rate) has been demonstrated in earlier studies. This study investigates the role of the distance of the precursor delivery point (nozzle) with respect to the substrate as well as the role of substrate tilting in minimizing the defects in graphene coatings developed on Ni and a Ni-Cu alloy by CVD, and durable corrosion resistance due to graphene coatings thus developed. Electrochemical impedance spectroscopy is employed to monitor the corrosion behavior of the coated samples during a 1008h immersion in a 0.1m NaCl solution. The findings reveal a significant reduction in corrosion rate of ≈88% and 98% for graphene-coated Ni and Ni-Cu alloys, respectively, compared to their uncoated counterparts.

Smart Monitoring Method for Land-Based Sources of Marine Outfalls Based on an Improved YOLOv8 Model

Land-based sources of marine outfalls are a major source of marine pollution. The monitoring of land-based sources of marine outfalls is an important means for marine environmental protection and governance. Traditional on-site manual monitoring methods are inefficient, expensive, and constrained by geographic conditions. Satellite remote sensing spectral analysis methods can only identify pollutant plumes and are affected by discharge timing and cloud/fog interference. Therefore, we propose a smart monitoring method for land-based sources of marine outfalls based on an improved YOLOv8 model, using unmanned aerial vehicles (UAVs). This method can accurately identify and classify marine outfalls, offering high practical application value. Inspired by the sparse sampling method in compressed sensing, we incorporated a multi-scale dilated attention mechanism into the model and integrated dynamic snake convolutions into the C2f module. This approach enhanced the model’s detection capability for occluded and complex-feature targets while constraining the increase in computational load. Additionally, we proposed a new loss calculation method by combining Inner-IoU (Intersection over Union) and MPDIoU (IoU with Minimum Points Distance), which further improved the model’s regression speed and its ability to predict multi-scale targets. The final experimental results show that the improved model achieved an mAP50 (mean Average Precision at 50) of 87.0%, representing a 3.4% increase from the original model, effectively enabling the smart monitoring of land-based marine discharge outlets.

Analysis of the accuracy of the inverse marching method used to determine thermal stresses in cylindrical pressure components

This paper analyses the inverse marching method used to determine the thermal stresses on the inner surface of a thick-walled cylindrical element not weakened by holes in the transient state. The heat conduction problem was considered one-dimensional, i.e. it was assumed that heat is transferred only in the radial direction. The method is based on measuring the temperature inside the pipeline wall at a single point and assuming that the pipeline is thermally insulated. The paper undertook an evaluation of the influence of the measuring point's distance from the inner surface, the number of control volumes into which the inverted area was divided and the length of the time step on the accuracy of the calculated temper-ature, heat transfer coefficient and thermal stresses on the inner surface of the pressure element. Verification was performed by comparing the calculation results obtained from the direct analytical method perturbed by random errors with those obtained from the numerical inverse step method.

МЕТОД МОНТЕ-КАРЛО ДЛЯ РАСЧЕТА ХАРАКТЕРИСТИК СВЕТОВЫХ ПОЛЕЙ В МОРСКОЙ ВОДЕ

The development of numerical methods for solving the integro-differential radiation transfer equation remains a relevant task. Among them, we can highlight the Monte Carlo method, which is in demand in various niches of modern ocean optics. The purpose of this work is a clear and concise presentation of the basics of the forward Monte Carlo method of light fields modeling in seawater, accompanied by a detailed description of its software implementation. The basics of the method are described, the procedures for choosing the type of interaction, the mean free path and the direction of photon motion are described. A simple case is considered, corresponding to an infinitely distant point source of unpolarized light, the absence of atmospheric influence, a smooth air-seawater interface, and the absence of stratification of inherent optical properties. In this case, realistic values of the absorption and scattering coefficients were used, calculated in accordance with the Case 1 model for a chlorophyll concentration of 1 μg/L, and a strongly elongated Henyey-Greenstein phase function with the parameter g = 0.95. The Fresnel reflection of light from the air-seawater interface was taken into account. The relative errors in the values of the diffuse attenuation coefficient for downward irradiance K d and the diffuse reflectance R, calculated in the spectral range of 400–700 nm using 106 photons, in comparison with the HydroLight results were 1.5 % and 0.4 %, respectively. Spectral calculation on one core of a 2017 Intel Core i5-8250U mobile processor in MATLAB takes 6 minutes. An assessment of the choice of the optimal number of photons required to obtain the desired quantities with a given accuracy was made. The implemented method is useful for becoming familiar with the basic principles used to numerically solve the radiative transfer equation in seawater using statistical methods and is used in the “Ocean Optics” course, taught by the author to 4th year students of the Department of Thermohydromechanics of the Ocean at MIPT.

A filter calibration method for laser-scanned weld toe geometries

The scanning of weld seams can be used to evaluate the local weld toe geometry for fatigue assessments. Laser scanned weld seam profiles often contain noise which complicates the accurate measurement of the weld toe geometry. For that reason, filtering of the scanned data is necessary. The issue at hand is that a filtering method can significantly affect the measurement results. Therefore, a calibration of the filter input parameters is needed. In this study, a calibration method for filtered laser-scanned weld profiles is presented by using artificial weld toe geometries. The adjustment of different filter functions is achieved by using an optimization method on predefined weld toes with an artificial noise. The resulting input data for the filter functions is tested on a real specimen to verify the method. Through the calibration method it is possible to achieve satisfactory measurement results with precisely set input parameters for the filter functions. The most suitable filter functions for the measurement of the weld toe are the Gaussian and the Lowpass filter. Both functions are adequate as a universally applicable filter. For the evaluation of the measurement results of the radii and angles, a tolerance range is introduced, which is defined by the theoretically minimum measurable radii and angles. Using an adjusted Lowpass filter and a point distance of 0.07 mm set by the laser scanner, a measurement within the tolerance range of 0.2 mm is achievable for the weld toe radius. For the weld toe angle, the tolerance range of 1.5° is achieved for the majority of measurements.

Air effects on solitary waves impinging and overtopping an impermeable seawall

Considering the presence or absence of air, the solitary waves impinging on an impermeable structure were numerically investigated to discuss solitary wave characteristics such as surface elevation, pressure, vorticity, eddy viscosity, and force on structures under different initial conditions. The volume of fluid (VOF) model (Wu, 2004) was employed to simulate solitary waves impinging an impermeable structure. Besides, the LES turbulent module was utilized to capture the turbulent phenomenon generated by wave-structure interactions, particularly after wave breaking. The present model was verified by comparing the simulation results with the experimental data (Hsiao and Lin, 2010). The findings demonstrated that the model more accurately represents tsunami wave behavior compared to simulations that did not account for air, especially in cases where wave breaking occurs prior to overtopping, entraining more air into the water. Detailed comparisons between simulations with and without air presence reveal that simulations excluding air exhibited delayed wave breaking, a more distant breaking point, and the absence of an air pocket behind the seawall. Despite these differences, the simulations showed excellent agreement with the laboratory results. Moreover, it was shown that the presence of air dissipated part of wave energy. Earlier the stage of waves broke, the more air was entrained, making the effects of air on the waves more noticeable. The highest fluid velocity was observed at the breaking point in front of the structure, while the strongest forces were experienced by the structure at the breaking location behind it.

Segmentation-Based Detection for Luffa Seedling Grading Using the Seg-FL Model

This study addresses the issue of inaccurate and error-prone grading judgments in luffa plug seedlings. A new Seg-FL seedling segmentation model is proposed as an extension of the YOLOv5s-Seg model. The small leaves of early-stage luffa seedlings are liable to be mistaken for impurities in the plug trays. To address this issue, cross-scale connections and weighted feature fusion are introduced in order to integrate feature information from different levels, thereby improving the recognition and segmentation accuracy of seedlings or details by refining the PANet structure. To address the ambiguity of seedling edge information during segmentation, an efficient channel attention module is incorporated to enhance the network’s focus on seedling edge information and suppress irrelevant features, thus sharpening the model’s focus on luffa seedlings. By optimizing the CIoU loss function, the calculation of overlapping areas, center point distances, and aspect ratios between predicted and ground truth boxes is preserved, thereby accelerating the convergence process and reducing the computational resource requirements on edge devices. The experimental results demonstrate that the proposed model attains a mean average precision of 97.03% on a self-compiled luffa plug seedling dataset, representing a 6.23 percentage point improvement over the original YOLOv5s-Seg. Furthermore, compared to the YOLACT++, FCN, and Mask R-CNN segmentation models, the improved model displays increases in mAP@0.5 of 12.93%, 13.73%, and 10.53%, respectively, and improvements in precision of 15.73%, 16.93%, and 13.33%, respectively. This research not only validates the viability of the enhanced model for luffa seedling grading but also provides tangible technical support for the automation of grading in agricultural production.

A Self-Calibration Method for Robot End-Effector Using Spherical Constraints

A self-calibration method utilizing spherical constraints is proposed for calibration of robot end-effectors. The method establishes a mathematical model to account for both the geometric errors of the robot and the deformation errors of the end-effector. A nonlinear least-squares parameter identification technique based on spherical constraints is employed to achieve autonomous calibration of the end-effector. Contrasted with methodologies relying on point plane or distance constraints, this novel technique delivers superior positioning accuracy, streamlined operational procedures and enhanced efficiency. Both simulation and experimental validation confirm that the self-calibration method using spherical constraints improves the positioning accuracy of the robot end-effector from 3 mm to 0.3 mm, showing the effectiveness of the method.

Cephalometric variation of vertical dimension in patients treated with hyrax-type and McNamara-type rapid palatal expander. Study on latero-lateral teleradiography

To analyze changes in the vertical dimension of the lower third of the face studied on teleradiograph in L-L following rapid palatal expansion achieved by four-band Hyrax-type rapid palatal expander (RPE) and by acrylic-bonded McNamara RPE to identify the most appropriate therapeutic choice based on the patient's facial growth pattern. 30 patients were selected, of whom 20 (9 males and 11 females with a mean age of 8.285 ± 1.216) were treated using McNamara RPE (Group D), and 10 (6 males and 4 females with a mean age of 8.562 ± 1.152) were treated with Hyrax RPE (Group B), for an average treatment time T0-T1 of 0.96 ± 0.501 years. According to Tweed and Ricketts, Cephalometric tracings obtained from lateral cephalograms at T0 and T1 were analyzed to study changes in the vertical dimension of the lower third of the face. For this purpose, 6 cephalometric landmarks were considered: convexity (distance of point A to the NPg plane), Ricketts’ total face height (angle between the NBa plane and Xi-Pm plane), lower face height (angle between the ANS and Xi-Pm planes), facial axis (angle between the NaBa plane and PtGn plane), ANB angle, and FMA angle. No statistically significant differences were found between the measurements at T0 and T1 for any of the 6 cephalometric measurements considered, neither within the single groups nor when comparing the two. However, a greater increasing trend was found for some variables between the two groups, such as Ricketts' total facial height, lower facial height, and FMA angle, although not statistically significant. Hyrax and McNamara's RPEs have provided minimal changes in the vertical component during palatal expansion treatment, thus demonstrating their ability to preserve the vertical dimension of the face. Therefore, there are no contraindications for using either appliance for patients with dolichofacial growth patterns.

Two-Staged Technology for CoCr Stent Production by SLM.

Additive manufacturing of porous materials with a specific macrostructure and tunable mechanical properties is a state-of-the-art area of material science. Additive technologies are widely used in industry due to numerous advantages, including automation, reproducibility, and freedom of design. Selective laser melting (SLM) is one of the advanced techniques among 3D fabrication methods. It is widely used to produce various medical implants and devices including stents. It should be noticed that there is a lack of information on its application in stent production. The paper presents the technological aspects of CoCr stent SLM fabrication, including design of stents and development of regimes for their manufacturing. Physical, chemical, and technological properties of CoCr powder were initially determined. Parametric design of mesh stent models was adopted. A two-stage approach was developed to ensure dimensional accuracy and quality of stents. The first stage involves a development of the single-track fusion process. The second stage includes the stent manufacturing according to determined technological regimes. The single-track fusion process was simulated to assign laser synthesis parameters for stent fabrication. Melting bath temperature and laser regimes providing such conditions were determined. Twenty-seven SLM manufacturing regimes were realized. Dependence of single-tracks width and height on the laser power, exposition time, and point distance was revealed. The qualitative characteristics of tracks imitating the geometry of the stent struts as well as favorable and unfavorable fusion regimes were determined. The results of surface roughness regulating of the stents' structural elements by various methods were analyzed. Thus, this two-staged approach can be considered as a fundamental approach for CoCr stent SLM fabrication.

Numerical simulation of electromagnetic-wave interference induced by ionization-front of millimeter-wave discharge at subcritical conditions and application to discharge structure identification

A standing wave induced in front of the ionization-front of a millimeter-wave discharge was numerically investigated to develop an interferometric discharge structure identification method. The time-varying waveform of the standing-wave intensity obtained at a distant observation point was smooth when a continuous comb-shaped structure was formed, whereas it was noisy with high-frequency components when a discrete structure was formed. The peak frequency of the Fourier spectrum of the time-varying waveform was proportional to the ionization-front propagation speed. The rapid time-variation of the waveform was caused by an increase in millimeter-wave absorption in a new plasma spot formation in the discrete structure. The results suggest that discharge structure identification, measurement of ionization-front propagation, and timing of plasma spot formation can be conducted experimentally without using a high-speed camera.

Hair-YOLO: a hair follicle detection model based on YOLOv8

Abstract Hair follicle detection technology has developed rapidly in recent years. Traditional manual detection methods are labor-intensive and inefficient. To address this problem, we propose a real-time hair follicle detection model called Hair-YOLO based on YOLOv8. This model focuses on accurately identifying the number of hairs within each follicle, providing precise data that help doctors assess hair density and follicle health of patients. First of all, we incorporate the re-parameterization Ghost module into the backbone, reducing the parameters and computational load. Then, the deformable convolution v3 operator is integrated into the neck network, enhancing adaptation to follicle shapes. Next, we propose a novel multi-scale feature perception separated and enhancement attention (Multi-SEAM) module, building upon the SEAM module, to address complex scalp scenarios. Furthermore, we enhance bounding box regression by replacing the standard complete intersection over union loss, with a modified point distance intersection over union loss. Finally, we construct a new hair follicle dataset and use it for a comparative analysis of Hair-YOLO and established models. Our model shows excellent performance, with a 14.26% increase in mAP@0.5:0.95, a 2.98% increase in Recall, and a 4.31% increase in Precision compared to the baseline.