Image-based Approach Research Articles

Bei Mu Gua Lou San facilitates mucus expectoration by increasing surface area and hydration levels of airway mucus in an air-liquid-interface cell culture model of the respiratory epithelium

BackgroundBei Mu Gua Lou San (BMGLS) is an ancient formulation known for its moisturizing and expectorant properties, but the underlying mechanisms remain unknown. We investigated concentration-dependent effects of BMGLS on its rehydrating and mucus-modulating properties using an air-liquid-interface (ALI) cell culture model of the Calu-3 human bronchial epithelial cell line and primary normal human bronchial epithelial cells (NHBE), and specifically focused on quantity and composition of the two major mucosal proteins MUC5AC and MUC5B.MethodsALI cultures were treated with BMGLS at different concentrations over three weeks and evaluated by means of histology, immunostaining and electron microscopy. MUC5AC and MUC5B mRNA levels were assessed and quantified on protein level using an automated image-based approach. Additionally, expression levels of the major mucus-stimulating enzyme 15-lipoxygenase (ALOX15) were evaluated.ResultsBMGLS induced concentration-dependent morphological changes in NHBE but not Calu-3 ALI cultures that resulted in increased surface area via the formation of herein termed intra-epithelial structures (IES). While cellular rates of proliferation, apoptosis or degeneration remained unaffected, BMGLS caused swelling of mucosal granules, increased the area of secreted mucus, decreased muco-glycoprotein density, and dispensed MUC5AC. Additionally, BMGLS reduced expression levels of MUC5AC, MUC5B and the mucus-stimulating enzyme 15-lipoxygenase (ALOX15).ConclusionsOur studies suggest that BMGLS rehydrates airway mucus while stimulating mucus secretion by increasing surface areas and regulating goblet cell differentiation through modulating major mucus-stimulating pathways.

Unmasking air quality: A novel image-based approach to align public perception with pollution levels

In the quest to reconcile public perception of air pollution with scientific measurements, our study introduced a pioneering method involving a gradient boost-regression tree model integrating PM2.5 concentration, visibility, and image-based data. Traditional stationary monitoring often falls short of accurately capturing public air quality perceptions, prompting the need for alternative strategies. Leveraging an extensive dataset of over 20,000 public visibility perception evaluations and over 8,000 stationary images, our models effectively quantify diverse air quality perceptions. The predictive prowess of our models was validated by strong performance metrics for perceived visibility (R = 0.98, RMSE = 0.19), all-day PM2.5 concentrations (R: 0.77–0.78, RMSE: 8.31–9.40), and Central Weather Bureau visibility records (R = 0.82, RMSE = 9.00). Interestingly, image contrast and light intensity hold greater importance than scenery clarity in the visibility perception model. However, clarity is prioritized in PM2.5 and Central Weather Bureau models. Our research also unveiled spatial limitations in stationary monitoring and outlined the variations in predictive image features between near and far stations. Crucially, all models benefit from the characterization of atmospheric light sources through defogging techniques. The image-based insights highlight the disparity between public perception of air pollution and current policy implementation. In other words, policymakers should shift from solely emphasizing the reduction of PM2.5 levels to also incorporating the public's perception of visibility into their strategies. Our findings have broad implications for air quality evaluation, image mining in specific areas, and formulating air quality management strategies that account for public perception.

Tumor classification of gastrointestinal liver metastases using CT-based radiomics and deep learning

ObjectivesThe goal of this study is to demonstrate the performance of radiomics and CNN-based classifiers in determining the primary origin of gastrointestinal liver metastases for visually indistinguishable lesions.MethodsIn this retrospective, IRB-approved study, 31 pancreatic cancer patients with 861 lesions (median age [IQR]: 65.39 [56.87, 75.08], 48.4% male) and 47 colorectal cancer patients with 435 lesions (median age [IQR]: 65.79 [56.99, 74.62], 63.8% male) were enrolled. A pretrained nnU-Net performed automated segmentation of 1296 liver lesions. Radiomics features for each lesion were extracted using pyradiomics. The performance of several radiomics-based machine-learning classifiers was investigated for the lesions and compared to an image-based deep-learning approach using a DenseNet-121. The performance was evaluated by AUC/ROC analysis.ResultsThe radiomics-based K-nearest neighbor classifier showed the best performance on an independent test set with AUC values of 0.87 and an accuracy of 0.67. In comparison, the image-based DenseNet-121-classifier reached an AUC of 0.80 and an accuracy of 0.83.ConclusionsCT-based radiomics and deep learning can distinguish the etiology of liver metastases from gastrointestinal primary tumors. Compared to deep learning, radiomics based models showed a varying generalizability in distinguishing liver metastases from colorectal cancer and pancreatic adenocarcinoma.

An image-based approach for the mass flow measurement of plastic granules as an alternative solution to loss-in-weight feeding systems

Advanced loss-in-weight systems are widely used for feeding operations to achieve high performance in several plastics manufacturing processes. However, these devices are expensive and unsuitable for dosing small batches of materials. Moreover, custom solutions are needed for those applications where extruders are operated in flood-fed conditions. Therefore, an image-based system for the mass flow measurement of plastic granules was developed and discussed in this work. Experimental data were collected to evaluate the performance of the proposed method. This approach exhibited high accuracy (± 1%) when dosing different batches, proving the image-based system suitable for production.

A Review of Face Detection Methods Based on Feature Approach

Face detection technology is the first and essential step for facial-based analysis algorithms such as face recognition, face feature extraction, face alignment, face enhancement, and face parsing. That is besides serving other applications related to human intention and act analysis such as facial expression recognition, gender recognition, and age classification. Face detection is used to detect faces in digital images. It is a special case of object detection and can be used in many areas such as biometrics, security, law enforcement, entertainment, and personal safety. There are various methods proposed in the field of face detection, and they all compete to make it more advanced and accurate. These methods belong to two main approaches; feature-based approach and image-based approach. This paper reviews the face detection methods that belong to the feature-based approach. Moreover, their work concepts, strengths, and limitations are mentioned. This paper concentrates on the feature-based approach due to its simplicity and high applicability in real-time detection compared to the image-based approach. Index Terms— Face Detection Methods, Feature-Based Approach, Viola-Jones Algorithm, Deformable Template Matching.

Natural language processing deep learning models for the differential between high-grade gliomas and metastasis: what if the key is how we report them?

The differential between high-grade glioma (HGG) and metastasis remains challenging in common radiological practice. We compare different natural language processing (NLP)-based deep learning models to assist radiologists based on data contained in radiology reports. This retrospective study included 185 MRI reports between 2010 and 2022 from two different institutions. A total of 117 reports were used for the training and 21 were reserved for the validation set, while the rest were used as a test set. A comparison of the performance of different deep learning models for HGG and metastasis classification has been carried out. Specifically, Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (BiLSTM), a hybrid version of BiLSTM and CNN, and a radiology-specific Bidirectional Encoder Representations from Transformers (RadBERT) model were used. For the classification of MRI reports, the CNN network provided the best results among all tested, showing a macro-avg precision of 87.32%, a sensitivity of 87.45%, and an F1 score of 87.23%. In addition, our NLP algorithm detected keywords such as tumor, temporal, and lobe to positively classify a radiological report as HGG or metastasis group. A deep learning model based on CNN enables radiologists to discriminate between HGG and metastasis based on MRI reports with high-precision values. This approach should be considered an additional tool in diagnosing these central nervous system lesions. The use of our NLP model enables radiologists to differentiate between patients with high-grade glioma and metastasis based on their MRI reports and can be used as an additional tool to the conventional image-based approach for this challenging task. • Differential between high-grade glioma and metastasis is still challenging in common radiological practice. • Natural language processing (NLP)-based deep learning models can assist radiologists based on data contained in radiology reports. • We have developed and tested a natural language processing model for discriminating between high-grade glioma and metastasis based on MRI reports that show high precision for this task.

A Learnable Image-Based Load Signature Construction Approach in NILM for Appliances Identification

One of the tasks of Non-Intrusive Load Monitoring (NILM) is load identification, which aims to extract and classify altered electrical signals after switching events are detected. In this subtask, representative and distinguishable load signatures are es-sential. At present, the literature approach to characterize electri-cal appliances is mainly based on manual feature engineering. However, the performance of signatures obtained by this way is limited. In this paper, we propose a novel load signature construc-tion method utilizing deep learning techniques. Specifically, three learnable load signatures are presented such as Learnable Recur-rent Graph (LRG), Learnable Gramian Matrix (LGM) and Gen-erative Graph (GG). Furthermore, we test different frameworks for learning these signatures and conclude that Temporal Convo-lutional Networks (TCN) based on residual learning are more suit-able for this work than the other schemes mentioned. The results of experiment on the PLAID datasets with submetered and aggre-gated, WHITED dataset and LILAC dataset confirm that our method outperforms the voltage-current trajectory, Recursive Graph and Gramian Angular Field methods in multiple evalua-tion metrics.

Location monitoring approach of underground pipelines using time-sequential images

The location monitoring of underground pipelines is of utmost significance as it helps the effective management and maintenance of the pipelines, and facilitates the planning of nearby projects, preventing damage to the pipelines. However, currently there is a serious lack of data on the locations of underground pipelines. This paper proposes an image-based approach for monitoring the locations of underground pipelines by combing deep learning and visual-based reconstruction. The proposed approach can build the monitoring model for underground pipelines and characterize their locations through their centroid curve. Its advantages are: (1) simplicity: it only requires time-sequential images of the inner walls of underground pipelines; (2) clarity: the location model and the location curve of underground pipelines can be provided quickly; (3) robustness: it can cope with some existing problems in underground pipelines, such as light variations and small viewing angles. A lightweight approach for monitoring the locations of underground pipelines is achieved. The proposed approach’s effectiveness has been validated through laboratory simulation experiments, demonstrating accuracy at the millimeter level.

A Virtual, 3D Multimodal Approach to Victim and Crime Scene Reconstruction.

In the last two decades, forensic pathology and crime scene investigations have seen a rapid increase in examination tools due to the implementation of several imaging techniques, e.g., CT and MR scanning, surface scanning and photogrammetry. These tools encompass relatively simple visualization tools to powerful instruments for performing virtual 3D crime scene reconstructions. A multi-modality and multiscale approach to a crime scene, where 3D models of victims and the crime scene are combined, offers several advantages. A permanent documentation of all evidence in a single 3D environment can be used during the investigation phases (e.g., for testing hypotheses) or during the court procedures (e.g., to visualize the scene and the victim in a more intuitive manner). Advanced computational approaches to understand what might have happened during a crime can also be applied by, e.g., performing a virtual animation of the victim in the actual context, which can provide important information about possible dynamics during the event. Here, we present an overview of the different techniques and modalities used in forensic pathology in conjunction with crime scene investigations. Based on our experiences, the advantages and challenges of an image-based multi-modality approach will be discussed, including how their use may introduce new visualization modalities in court, e.g., virtual reality (VR) and 3D printing. Finally, considerations about future directions in research will be mentioned.

Towards image-based laryngeal videostroboscopy using deep learning-enabled compressed sensing

Laryngeal videostroboscopy is an audio-mediated imaging technique allowing the visualization of vocal fold oscillation behavior: the audio signal is used to determine the fundamental frequency F0, which represents the vocal fold oscillation frequency. Knowing F0 allows to trigger the strobe illumination unit to provide a still image or slow-motion view of the vocal fold oscillation. However, this procedure involves several hardware components, noisy audio signals, and a chain of complex, error-prone algorithms that have to be orchestrated. We hypothesize that endoscopic images suffice to determine F0 with a view towards providing an alternative, image-based approach for estimating F0 during laryngeal videoendoscopy. In this study, we show that we are able to predict the relative glottal opening state to create sample points on the glottal area waveform, an endoscopic image-derived signal capable of deriving F0. As imaging frame rates from ordinary endoscopic cameras do not fulfill the Shannon–Nyquist criterion, we solve this problem with compressed sensing. We developed and evaluated the proposed approach using high-speed videoendoscopy (HSV) to simulate different, realistic low frame rates that are similar to those used in videostroboscopy. We show that we are able to predict F0 with over 95% accuracy using at most 75 sample points of a 600 ms long footage. Using endoscopic images and our algorithm only, we showcase that we can achieve a stroboscopic effect. This shows, that our proposed method in combination with the developed algorithm may be considered in the future to be integrated into clinical videostroboscopy.

Virtual-specimen modeling of aggregate contact effects on asphalt concrete

Aggregate contacts significantly affect the mechanical behavior of asphalt concrete. However, there still lacks an effective way to account for them in numerical modeling. Therefore, to address this concern, a new virtual-specimen-based modeling approach was developed in this study by simplifying aggregate particles in asphalt concrete as spheres and incorporating aggregate contacts through Contact Region (CR) elements. The complex moduli of both gap-graded and dense-graded mixtures were predicted using this approach and compared with those predicted using the conventional image-based modeling approach and laboratory-measured values. The virtual-specimen modeling revealed that the CR in the gap-graded mixture with a higher proportion of large aggregates can better transmit load among aggregates than in the dense-graded mixture. Both modeling approaches were found to provide good prediction accuracy, but the stress distributions in the virtual-specimen models were more uniform and continuous, leading to better computational convergence and the possibility of nonlinear analysis of asphalt concrete.

Static Shape Control of Soft Continuum Robots Using Deep Visual Inverse Kinematic Models

Soft continuum robots are highly flexible and adaptable, making them ideal for unstructured environments such as the human body and agriculture. However, their high compliance and maneuverability make them difficult to model, sense, and control. Current control strategies focus on Cartesian space control of the end-effector, but few works have explored full-body control. This study presents a novel image-based deep learning approach for closed-loop kinematic shape control of soft continuum robots. The method combines a local inverse kinematics formulation in the image space with deep convolutional neural networks for accurate shape control that is robust to feedback noise and mechanical changes in the continuum arm. The shape controller is fast and straightforward to implement; it takes only a few hours to generate training data, train the network, and deploy, requiring only a web camera for feedback. This method offers an intuitive and user-friendly way to control the robot's 3-D shape and configuration through teleoperation using only 2-D hand-drawn images of the desired target state without the need for further user instruction or consideration of the robot's kinematics.

Experimental investigation of electrode wear assessment in the EDM process using image processing technique

Abstract One of the widespread non-traditional and non-contact machining processes is the electrical discharge machining (EDM) process, which depends on removing material from the surface by different repeat electrical discharges between a machined workpiece and a cutting tool called the electrode. The electrode wear in the EDM process is one of the parameters of significant technological interest. This article proposed a reliable image-based system that automatically detects and computes electrode wear in EDM of rectangular steel workpiece plate of AISI 314 using a copper electrode with a rectangular shape. MATLAB environment and image processing toolbox are used to identify and manipulate the captured rectangular electrode image for rectangular electrode data extraction needed for the electrode wear identification process. The results show that the proposed image-based approach, using a non-contact measuring system and relatively inexpensive equipment, is suitable with good accuracy and efficiency for measuring and testing electrode wear.

Real-time in vivo dose measurement using ruby-based fibre optic dosimetry during internal radiation therapy

In vivo dosimetry (IVD) in a commonly used liver cancer treatment of selective internal radiation therapy (SIRT) has been done based on the post-treatment image-based dosimetry approach. Real-time IVD is necessary to verify the dose delivery and detect errors during the treatment for better patient outcomes. This study aims to develop a fibre optic dosimeter (FOD) for in vivo real-time dose rate measurement during internal beta radiation therapy, e.g., SIRT. A ruby fibre optic probe was prepared and studied the radioluminescence (RL) characteristics, including its major challenge of stem effect arising from Cherenkov radiation and luminescence from the irradiated fibre. The stem signal was suppressed adequately using the stem removal technique of optical filtering, and only 2.3 ± 1.1% stem signal was contributed to the measured RL signal. A linear dose rate response was observed during the exposure of the ruby probe to varying dose rates using a 6 MeV electron beam and a positron-emitting radionuclide fluorine-18. The ruby exhibited a temporally non-constant RL signal, which increased the RL signal by 0.84 ± 0.29 counts/sec2 during the irradiation of the maximum dose rate used in this study of 9 Gy/min for 2 min. The ability of ruby FOD to measure the absolute dose rate with sufficient stem effect suppression and the linear RL dose rate response indicates its suitability for real-time IVD during internal beta radiation therapy. Future work will investigate the time-dependent RL characteristic of ruby and validate post-treatment image-based dosimetry using ruby-based FOD.

Variation analysis of photonic lantern mode control ability under mode oscillation: a new approach.

In this Letter, we present a novel, to the best of our knowledge, image-based approach to analyze the mode control ability of a photonic lantern employed in diode laser beam combining, aiming to achieve a stable beam output. The proposed method is founded on theories of power flow and mode coupling and is validated through experiments. The findings demonstrate that the analysis of the beam combining process is highly reliable when the main mode component of the output light is the fundamental mode. Moreover, it is experimentally demonstrated that the mode control performance of the photonic lantern significantly influences the beam combining loss and the fundamental mode purity. In the essence of the variation-based analysis, a key advantage of the proposed method is its applicability even in the situation of a poor combined beam stability. The experiment only requires the collection of the far-field light images of the photonic lantern to characterize the model control ability, achieving an accuracy greater than 98%.

Weld defect identification and characterization in radiographic images using deep learning: Review

Defects in the welds degrade the quality of the weld. Weld defect identification is a challenging task in the industry because of the wide range of weld imperfections. Weld defect detection using radiographic images is an effective technique for achieving good weld quality in shipbuilding and aerospace applications. Foreign inclusions, cracks and pores are examples of welding joint imperfections. Several appropriate computer-based image processing techniques have made the detection of weld defects possible. It is challenging because weld imperfection can show various sizes, shapes, contrasts and locations in radiography images. The accuracy of this inspection process is more dependent on various external factors and is also time-consuming. Automatic weld defect detection is needed by analyzing the images obtained directly from digital radiographic systems. This paper uses a unique image-based approach to a small batch of x-ray imaging datasets to investigate a potential solution for weld defect identification. This article compares a deep learning network’s performance for various parameter and hyper-parameter combinations. Also it compares the traditional approaches of defect detection using manual inspection method, feature-based defect identification, and finally deep - learning based approach on several types of weld defects in various industrial applications. This comparative analysis concludes that deep learning-based approaches have achieved more accuracy as compared to conventional techniques. This research paper also highlights a few challenges and future directions in welding area.

인공 신경망 예측기를 이용한 새로운 데이터 모델 기반 PV 어레이 고장 진단용 기준 신호 발생 방법

The power supply part of a PV system is called a PV array which is composed of many PV strings connected in parallel. A PV-string which is a set of PV modules is made by serial connection of many PV modules. PV modules presents irregularity of ouput power which is generated due to various factors such as average performance degradation of about 0.923% to 1.54% according to various researches and born power variations over production stages. Power degradation and born power variation are due to various factors such as material interactions(connection state of connectors between PV cells, corrosion, browning of encapsulation materials ...etc) and environment factors such as shading and soiling which refers to the accumulation of snow, dirt, leaves and bird droppings on PV modules. Such various minor factors can make side effects in safety side and economic one. Therefore several methodology for diagnosis of PV-array have been developed, which are classified into three types – image-based diagnosis approach, model-based approach and data-driven approach. In this paper, a new data-based approach(called data model) with good failure diagnosis reliability and economy is proposed, and verified by simulation using Python, published data and reasonable data generation. Based on the bathtub failure rate function, the fault diagnosis model requires a well-functioning predictor to generate a reference signal which evaluates the output characteristics of PV array under instantaneously varying environmental condition such as solar irradiance, temperature, etc. To implement a new data-based approach, an Artificial Neural Network (ANN) based predictor is applied as a reference generator for PV array’s fault diagnosis. r² value and the RMSE(Root Mean Square Error) are used to evaluate how far away individual learned predictions are from the actual measured values. Based on the reliability obtained from the learning result for a specific PV string, it is confirmed through pair comparison analysis(called t-test) that the learning result(called clustering possibility) between PV strings under different installation conditions is also reliable.

Machine Learning for Fog-and-Low-Stratus Nowcasting from Meteosat SEVIRI Satellite Images

Fog and low stratus (FLS) are meteorological phenomena that have a significant impact on all ways of transportation and public safety. Due to their similarity, they are often grouped together as a single category when viewed from a satellite perspective. The early detection of these phenomena is crucial to reduce the negative effects that they can cause. This paper presents an image-based approach for the short-term nighttime forecasting of FLS during the next 5 h over Morocco, based on geostationary satellite observations (Meteosat SEVIRI). To achieve this, a dataset of hourly night microphysics RGB product was generated from native files covering the nighttime cold season (October to April) of the 5-year period (2016–2020). Two optical flow techniques (sparse and dense) and three deep learning techniques (CNN, Unet and ConvLSTM) were used, and the performance of the developed models was assessed using mean squared error (MSE) and structural similarity index measure (SSIM) metrics. Hourly observations from Meteorological Aviation Routine Weather Reports (METAR) over Morocco were used to qualitatively compare the FLS existence in METAR, where it is also shown by the RGB product. Results analysis show that deep learning techniques outperform the traditional optical flow method with SSIM and MSE of about 0.6 and 0.3, respectively. Deep learning techniques show promising results during the first three hours. However, their performance is highly dependent on the number of filters and the computing resources, while sparse optical flow is found to be very sensitive to mask definition on the target phenomenon.

(298) MRI Measurement of Perineal and Pelvic Muscles as a Novel Imaging Marker for Hypogonadism

Abstract Introduction Low serum testosterone (T) in the presence of clinical symptoms of hypogonadism is essential in determining who may benefit from T replacement therapy. However, clinical symptoms associated with hypogonadism are non-specific and challenging to discriminate from other etiologies. While low bone mineral density is an objective primary end-organ effect of low testosterone, DEXA scans are not recommended by national guidelines due to poor sensitivity and specificity, and the limitation that these changes likely manifest over a prolonged period of time. Notably, prior studies have demonstrated that perineal ultrasound of the bulbocavernosus muscle (BCM) can be an effective surrogate to evaluate end-organ activity of hypogonadism. However, it is unclear if low BCM area is secondary to or rather correlated with hypogonadism, and longitudinal studies assessing men in both the eugonadal and hypogonadal state are lacking. Objective To report an initial experience with measuring the androgen-sensitive BCM, ischiocavernosus muscle (ICM), and levator ani muscles (LAM) on MRI in men before and after androgen deprivation therapy (ADT) as a novel imaging marker for end-organ effects of hypogonadism. Methods Data was collected from treatment-naive patients with intermediate or high-risk localized prostate cancer enrolled in a clinical trial involving six months of neoadjuvant treatment with goserelin and enzalutamide prior to prostatectomy. Patients underwent 3-Tesla multiparametric MRI of the prostate prior to treatment and again immediately prior to prostatectomy. An expert genitourinary trained radiologist retrospectively assessed MR images and measured width of the bilateral BCM, ICM, and LAM on coronal T2W MRI (Figure 1). Paired t-tests were performed to evaluate changes in muscle width before and after ADT. Linear regression was performed to evaluate the relationship between change in total testosterone levels and change in muscle width. The lower detection limit of testosterone was 20 ng/dL, and T less than 20 ng/dL was set to 0 for statistical analysis. Results 38 patients with pre- and post-treatment MRIs were identified with a median time between scans of 6.85 months. Median age at start of trial was 62.5 years (IQR: 58-69). 28/38 patients had documented AUA/IPSS and SHIM questionnaires with median scores of 8 and 19.5, respectively. Mean total testosterone prior to androgen deprivation was 289.5 ng/dL (SD= 121.7). 34/38 patients reached castrate level (T less than 20 ng/dL) after treatment completion, with mean testosterone 4.7 ng/dL (SD= 15.4). Muscle width was significantly reduced in all three muscle groups following ADT (Table 1). Linear regression found that decreases in testosterone significantly predicted decreases in combined perineal muscle (BCM + ICM) width (β = 0.0056, p = 0.032). Conclusions Decreases in testosterone to castrate levels led to significant and rapid decreases in sizes of both perineal and pelvic muscles. This shows initial plausibility in using an image-based approach to assess a short- to intermediate-term objective marker of end-organ effects of hypogonadism. Future prospective studies with larger samples and refinement of MRI sequences may lead to a useful clinical tool in the diagnosis and clinical workup of hypogonadism, and help clinicians better understand who would benefit from testosterone supplementation. Disclosure No

An Efficient Adaptive Noise Removal Filter on Range Images for LiDAR Point Clouds

Light Detection and Ranging (LiDAR) is a critical sensor for autonomous vehicle systems, providing high-resolution distance measurements in real-time. However, adverse weather conditions such as snow, rain, fog, and sun glare can affect LiDAR performance, requiring data preprocessing. This paper proposes a novel approach, the Adaptive Outlier Removal filter on range Image (AORI), which combines a projection image from LiDAR point clouds with an adaptive outlier removal filter to remove snow particles. Our research aims to analyze the characteristics of LiDAR and propose an image-based approach derived from LiDAR data that addresses the limitations of previous studies, particularly in improving the efficiency of nearest neighbor point search. Our proposed method achieves outstanding performance in both accuracy (>96%) and processing speed (0.26 s per frame) for autonomous driving systems under harsh weather from raw LiDAR point clouds in the Winter Adverse Driving dataset (WADS). Notably, AORI outperforms state-of-the-art filters by achieving a 6.6% higher F1 score and 0.7% higher accuracy. Although our method has a lower recall than state-of-the-art methods, it achieves a good balance between retaining object points and filter noise points from LiDAR, indicating its promise for snow removal in adverse weather conditions.