Image Texture Features Research Articles

Research on Target Image Classification in Low-Light Night Vision.

In extremely dark conditions, low-light imaging may offer spectators a rich visual experience, which is important for both military and civic applications. However, the images taken in ultra-micro light environments usually have inherent defects such as extremely low brightness and contrast, a high noise level, and serious loss of scene details and colors, which leads to great challenges in the research of low-light image and object detection and classification. The low-light night vision image used as the study object in this work has an excessively dim overall picture and very little information about the screen's features. Three algorithms, HE, AHE, and CLAHE, were used to enhance and highlight the image. The effectiveness of these image enhancement methods is evaluated using metrics such as the peak signal-to-noise ratio and mean square error, and CLAHE was selected after comparison. The target image includes vehicles, people, license plates, and objects. The gray-level co-occurrence matrix (GLCM) was used to extract the texture features of the enhanced images, and the extracted image texture features were used as input to construct a backpropagation (BP) neural network classification model. Then, low-light image classification models were developed based on VGG16 and ResNet50 convolutional neural networks combined with low-light image enhancement algorithms. The experimental results show that the overall classification accuracy of the VGG16 convolutional neural network model is 92.1%. Compared with the BP and ResNet50 neural network models, the classification accuracy was increased by 4.5% and 2.3%, respectively, demonstrating its effectiveness in classifying low-light night vision targets.

7847 Radiomic Ultrasound Checkup of Thyroid Nodules Before Radiodiodine Therapy

Abstract Disclosure: A.A. Trukhin: Grant Recipient; Self; Russian Science Foundation (project N 22-15-00135). A.V. Manaev: Grant Recipient; Self; Russian Science Foundation (project N 22-15-00135). S.M. Zakharova: Grant Recipient; Self; Russian Science Foundation (project N 22-15-00135). M.S. Sheremeta: None. E.A. Troshina: Grant Recipient; Self; Russian Science Foundation (project N 22-15-00135). Background: Over 83 years radioiodine therapy (RIT) of autonomous thyroid nodules still have uncovered pages, according to its complexity and thyroid nodules variety. There are two main criteria to allow RIT – no malignancy and presence of iodine uptake, the last one could be checked by Tc-99m-pertachnetate or I-123, I-131 iodine scintigraphy. Case with thyroid nodules first undergo ultrasound (US) TI-RADS classification, then fine needle aspiration stratification according to the Bethesda system. Publications applied such algorithm shows 75-85% accuracy in malignancy determination, that in 15-25% can lead to improper RIT application. Aim: Method development for thyroid nodules ultrasound checkup before I-131 administration and I-131 24h uptake prognosis. Methods: Study includes 150 nodules (65 benign, 60 malignant, 25 autonomous benign thyroid nodules). Both US longitudinal and transversal thyroid nodule projections were obtained via GE Voluson E8 (36% benign cases / 27% malignant cases) and GE Logiq E (64% benign cases / 73% malignant cases) in cinematic loop regime. Autonomous thyroid nodules underwent I-131 pharmacokinetics examination with 24h uptake determination and RIT. Statistical textural features were obtained through spatial adjacency matrix. Results: Analysis of 16 statistical textural features shows weak correlation between 11 statistical textural features E (Energy), S (Entropy), LU (Local homogeneity), MP (Maximum probability), TR (Matrix trace), CORR (Correlation), IMC (information measure of correlation), DV (Difference variance), DE (Difference entropy), SE (Sum entropy), SA (Average value of the sum). ROC-analysis using 11 statistical textural features shows the best model for benign/malignance determination with AUC = 0.83, Se = 74%, Sp = 73%, PPV = 75%, NPV = 72%, Acc = 74%. I-131 pharmacokinetics examination shows strong correlation SA to I-131 24h uptake. Conclusions: The performed study demonstrates the possibility of using textural statistical features of ultrasound images as instrument for thyroid nodules ultrasound checkup before I-131 administration, for extra FNA prescription and prognosis the I-131 24h important in exact therapeutic I-131 activity calculation. Presentation: 6/2/2024

Radiomic Texture Features in CT Images of Kidneys in Ventilated Deceased Donors Predict Delayed Graft Function

Radiomic Texture Features in CT Images of Kidneys in Ventilated Deceased Donors Predict Delayed Graft Function

A computer vision system and machine learning algorithms for prediction of physicochemical changes and classification of coated sweet cherry

The current research utilized visual characteristics obtained from RGB images and qualitative characteristics to investigate changes in surface defects, predict physical and chemical characteristics, and classify sweet cherries during storage. It was achieved with the help of ANN (Artificial Neural Network) and ANFIS (Adaptive Neuro-Fuzzy Inference System) models. The ANN used in this study was a Multilayer Perceptron (MLP) with SigmoidAxon and TanhAxon threshold functions, trained with the Momentum training function. Additionally, ANFIS with a Mamdani system and Triangle, Gauss, and Trapezoidal membership functions, was employed to predict sweet cherries' physical and chemical properties and their quality classification. Both models incorporate four algorithms. Additionally, the algorithms use color statistical features and color texture features combined with physical and chemical properties, including weight loss, firmness, titratable acidity, and total anthocyanin content. The image color and texture characteristics were used by ANN and ANFIS models to predict physical and chemical properties with high accuracy. ANN and ANFIS models accurately estimate sweet cherry quality grades in all four algorithms with over 90% accuracy. According to the findings, the ANN and ANFIS models have demonstrated satisfactory performance in the qualitative classification and prediction of sweet cherries' physical and chemical properties.

Research on rapid prediction method of laser cladding deposited layer state based on molten pool texture sequence

A texture sequence feature combining texture statistical characteristics with time series is proposed for rapid prediction of the state of laser cladding deposited layer. Texture features in the molten pool images are extracted using high-order dense texture descriptor operators and transformed into probability sequences with the aid of multi-cell histograms. Texture features are screened based on the types of points of interest to improve the sequence, enhancing the prediction accuracy and speed for deposited layer. For different types of points of interest, the optimal texture sequence obtained achieves an improvement of more than 6% in prediction accuracy for deposited layer state compared to the full-texture sequence. The proposed texture sequence features, when compared with molten pool image features, exhibit an average improvement in prediction speed of more than 17 times across different deep learning models. Overall, the results indicate that this method significantly improves prediction speed while maintaining accuracy, satisfying the requirements of high-speed monitoring in laser cladding.

Image texture analysis of pellets made of lignocellulosic materials

This experiment aimed to find the relation between image texture features of pellets made of various lignocellulosic materials (wood, wheat straw, hay, giant miscanthus, prairie spartina, and giant knotweed) and their physico-mechanical properties (density, compressive energy, maximum compressive strength, modulus of elasticity). Using the Kruskal-Wallis's test, the effect of materials on these properties was examined. Texture features were derived from the grey-level co-occurrence matrix, grey-level run-length matrix, absolute gradient matrix, autoregressive model, and wavelet decomposition, resulting in 86 features, later reduced to 8 factors via explanatory factor analysis. These factors were used as predictors in regression models for physico-mechanical properties. The models for modulus of elasticity achieved R2adj values of 0.91–0.99 (except for hay and wood), compressive stress models achieved 0.65–0.99 (excluding hay and wood), compressive energy models ranged from 0.60 to 0.97 (excluding hay), and density models ranged from 0.56 to 0.97 (excluding wood). The study confirmed a significant correlation between material type, texture parameters and compression resistance, suggesting this method could monitor pellet quality in production.

Classification method of seabed sonar image substrate based on ELM-AdaBoost

The results of sediment classification are affected by the measurement environment. The water depth has a significant impact on the sound propagation, and the acoustic characteristics will be different under different water depths. In addition, seabed topography, sediment types and their distribution, acoustic frequency, equipment types and settings, and suspended substances in water will all affect the original sonar data. In order to obtain accurate classification results, a seabed sonar image classification method based on Elm AdaBoost is proposed. The original sound intensity data is compensated for gain and the anomaly is eliminated. After preprocessing the multi beam sonar data, the image texture feature is extracted from the gray level co-occurrence matrix as the feature vector for classification, and the elm classifier is constructed. The elm is enhanced by AdaBoost. The calculated feature is input into the elm AdaBoost network for sediment classification, and the result is output. Simulation results show that the proposed method effectively improves the accuracy of sediment classification, and verifies the feasibility of this method.

Classification and identification of pest, diseases and nutrient deficiency in paddy using layer based EMD phase features with decision tree

Pest attack, disease incidence, and nutrient deficiency are the major factors limiting the yield of paddy. Therefore, the paper proposes a classification system for the identification of pest, disease, and nutrient deficiency classes. This approach initially preprocesses leaf images using entropy filtering followed by a leaf segmentation process. Multiple layers are then constructed on the leaf image through which features are extracted. The Gray Level Co-occurrence Matrix (GLCM) algorithm and Principal Component Analysis (PCA) are used to extract the global texture features of the leaf image. A 1D-signal sequence is constructed on each layer, which is decomposed by the Empirical Mode Decomposition algorithm from which the phase features are estimated. The features are trained/classified using the decision tree classifiers that classify the pest attack, disease incidence, and nutrient deficiency categories. The proposed approach provides a precision, accuracy, specificity, sensitivity, and F1-score of 97 %, 97.88 %, 96.52 %, 96.7 %, and 96.7 % respectively.

Estimation of winter wheat LAI based on color indices and texture features of RGB images taken by UAV.

Leaf area index (LAI) is an important indicator for assessing plant growth and development, and is also closely related to photosynthesis in plants. The realization of rapid accurate estimation of crop LAI plays an important role in guiding farmland production. In study, the UAV-RGB technology was used to estimate LAI based on 65 winter wheat varieties at different fertility periods, the wheat varieties including farm varieties, main cultivars, new lines, core germplasm and foreign varieties. Color indices (CIs) and texture features were extracted from RGB images to determine their quantitative link to LAI. The results revealed that among the extracted image features, LAI exhibited a significant positive correlation with CIs (r = 0.801), whereas there was a significant negative correlation with texture features (r = -0.783). Furthermore, the visible atmospheric resistance index, the green-red vegetation index, the modified green-red vegetation index in the CIs, and the mean in the texture features demonstrated a strong correlation with the LAI with r > 0.8. With reference to the model input variables, the backpropagation neural network (BPNN) model of LAI based on the CIs and texture features (R2 = 0.730, RMSE = 0.691, RPD = 1.927) outperformed other models constructed by individual variables. This study offers a theoretical basis and technical reference for precise monitor on winter wheat LAI based on consumer-level UAVs. The BPNN model, incorporating CIs and texture features, proved to be superior in estimating LAI, and offered a reliable method for monitoring the growth of winter wheat. © 2024 Society of Chemical Industry.

Planetary gearbox fault classification based on tooth root strain and GAF pseudo images

Traditional signal processing methods based on acceleration signals can determine whether a fault has occurred in a planetary gearbox. However, acceleration signals are severely affected by interference, causing difficulties in fault identification. This study proposes a gear fault classification method based on root strain and pseudo images. Firstly, fiber optic sensors are employed to directly acquire strain data from the ring gear root. Next, the strain signals are preprocessed using resampling and a time-domain synchronous averaging algorithm. The processed signals are encoded into two-dimensional images using Gramian Angular Fields (GAF). Then, CN-EfficientNet with contrast learning is proposed to analyze and extract deeper fault features from the image texture features. In the classification experiments for different types of faults, the accuracy reached 96.84%. The results indicate that the method can effectively accomplish the task of fault classification in planetary gearboxes. Comparative experiments with other common classification models further indicate the superior performance of the proposed learning model. Visualization based on Grad-CAM provides interpretability for the fault recognition network’s results and reveals the underlying mechanism for its excellent classification performance.

Texture-based speciation of otitis media-related bacterial biofilms from optical coherence tomography images using supervised classification.

Otitis media (OM), a highly prevalent inflammatory middle-ear disease in children worldwide, is commonly caused by an infection, and can lead to antibiotic-resistant bacterial biofilms in recurrent/chronic OM cases. A biofilm related to OM typically contains one or multiple bacterial species. OCT has been used clinically to visualize the presence of bacterial biofilms in the middle ear. This study used OCT to compare microstructural image texture features from bacterial biofilms. The proposed method applied supervised machine-learning-based frameworks (SVM, random forest, and XGBoost) to classify multiple species bacterial biofilms from in vitro cultures and clinically-obtained in vivo images from human subjects. Our findings show that optimized SVM-RBF and XGBoost classifiers achieved more than 95% of AUC, detecting each biofilm class. These results demonstrate the potential for differentiating OM-causing bacterial biofilms through texture analysis of OCT images and a machine-learning framework, offering valuable insights for real-time in vivo characterization of ear infections.

The Prediction and Evaluation of Surface Quality during the Milling of Blade-Root Grooves Based on a Long Short-Term Memory Network and Signal Fusion.

The surface quality of milled blade-root grooves in industrial turbine blades significantly influences their mechanical properties. The surface texture reveals the interaction between the tool and the workpiece during the machining process, which plays a key role in determining the surface quality. In addition, there is a significant correlation between acoustic vibration signals and surface texture features. However, current research on surface quality is still relatively limited, and most considers only a single signal. In this paper, 160 sets of industrial field data were collected by multiple sensors to study the surface quality of a blade-root groove. A surface texture feature prediction method based on acoustic vibration signal fusion is proposed to evaluate the surface quality. Fast Fourier transform (FFT) is used to process the signal, and the clean and smooth features are extracted by combining wavelet denoising and multivariate smoothing denoising. At the same time, based on the gray-level co-occurrence matrix, the surface texture image features of different angles of the blade-root groove are extracted to describe the texture features. The fused acoustic vibration signal features are input, and the texture features are output to establish a texture feature prediction model. After predicting the texture features, the surface quality is evaluated by setting a threshold value. The threshold is selected based on all sample data, and the final judgment accuracy is 90%.

PAFE: A lightweight visualization-based fast malware classification method

With the development of automated malware toolkits, cybersecurity faces evolving threats. Although visualization-based malware analysis has proven to be an effective method, existing approaches struggle with challenging malware samples due to alterations in the texture features of binary images during the visualization preprocessing stage, resulting in poor performance. Furthermore, to enhance classification accuracy, existing methods sacrifice prediction time by designing deeper neural network architectures. This paper proposes PAFE, a lightweight and visualization-based rapid malware classification method. It addresses the issue of texture feature variations in preprocessing through pixel-filling techniques and applies data augmentation to overcome the challenges of class imbalance in small sample datasets. PAFE combines multi-scale feature fusion and a channel attention mechanism, enhancing feature expression through modular design. Extensive experimental results demonstrate that PAFE outperforms the current state-of-the-art methods in both efficiency and effectiveness for malware variant classification, achieving an accuracy rate of 99.25 % with a prediction time of 10.04 ms.

Examining feature extraction and classification modules in machine learning for diagnosis of low-dose computed tomographic screening-detected in vivo lesions.

Medical imaging-based machine learning (ML) for computer-aided diagnosis of in vivo lesions consists of two basic components or modules of (i)feature extraction from non-invasively acquired medical images and (ii)feature classification for prediction of malignancy of lesions detected or localized in the medical images. This study investigates their individual performances for diagnosis of low-dose computed tomography (CT) screening-detected lesions of pulmonary nodules and colorectal polyps. Three feature extraction methods were investigated. One uses the mathematical descriptor of gray-level co-occurrence image texture measure to extract the Haralick image texture features (HFs). One uses the convolutional neural network (CNN) architecture to extract deep learning (DL) image abstractive features (DFs). The third one uses the interactions between lesion tissues and X-ray energy of CT to extract tissue-energy specific characteristic features (TFs). All the above three categories of extracted features were classified by the random forest (RF) classifier with comparison to the DL-CNN method, which reads the images, extracts the DFs, and classifies the DFs in an end-to-end manner. The ML diagnosis of lesions or prediction of lesion malignancy was measured by the area under the receiver operating characteristic curve (AUC). Three lesion image datasets were used. The lesions' tissue pathological reports were used as the learning labels. Experiments on the three datasets produced AUC values of 0.724 to 0.878 for the HFs, 0.652 to 0.965 for the DFs, and 0.985 to 0.996 for the TFs, compared to the DL-CNN of 0.694 to 0.964. These experimental outcomes indicate that the RF classifier performed comparably to the DL-CNN classification module and the extraction of tissue-energy specific characteristic features dramatically improved AUC value. The feature extraction module is more important than the feature classification module. Extraction of tissue-energy specific characteristic features is more important than extraction of image abstractive and characteristic features.

Classification methods for hyperspectral remote sensing images with weak texture features

In order to provide method support for remote sensing image classification of weak texture region and improve the classification effect, a remote sensing image classification model for weak texture region was constructed. Firstly, the Gray Level Co-occurrence Matrix (GLCM) was used to extract texture features, then the correlation between texture features was reduced by dimensionality reduction method, and finally the spectral features were fused to classify remote sensing images. In the experimental part, neural network classifier was used to classify images based on texture feature, spectral feature and spectral features combined with texture features, and the results ware compared with object-oriented classification method. The analysis of texture extraction in the study area shows that the texture feature image obtained in the 5 × 5 window has higher resolution than that in the 7 × 7 window, and more robust texture features can be obtained in the range of 4∼8 pixel pairs. When using neural network classifier for classification, compared with images based on spectral features and texture features, the overall classification accuracy (OA) of the proposed classification model is increased by 16.72% and 11.16%, respectively, and the Kappa coefficient is increased by 0.2824 and 0.1943, respectively. Compared with the object-oriented classification method, the overall classification accuracy of the proposed classification scheme is increased by 1.15%, and the Kappa coefficient is increased by 0.0191. The constructed model has good classification effect and high classification accuracy for remote sensing images of weak texture region, and can provide scientific reference for remote sensing image classification of weak texture region.

Texture differences of microchambers and macrochambers in heel pads between the elderly with and without diabetes

ObjectiveThis study aims to use the texture analysis of ultrasound images to distinguish the features of microchambers (a superficial thinner layer) and macrochambers (a deep thicker layer) in heel pads between the elderly with and without diabetes, so as to preliminarily explore whether texture analysis can identify the potential injury characteristics of deep tissue under the influence of diabetes before the obvious injury signs can be detected in clinical management. MethodsUltrasound images were obtained from the right heel (dominant leg) of eleven elderly people with diabetes (DM group) and eleven elderly people without diabetes (Non-DM group). The TekScan system was used to measure the peak plantar pressure (PPP) of each participant. Six gray-level co-occurrence matrix (GLCM) features including contrast, correlation, dissimilarity, energy, entropy, homogeneity were used to quantify texture changes in microchambers and macrochambers of heel pads. ResultsSignificant differences in GLCM features (correlation, energy and entropy) of macrochambers were found between the two groups, while no significant differences in all GLCM features of microchambers were found between the two groups. No significant differences in PPP and tissue thickness in the heel region were observed between the two groups. ConclusionsIn the elderly with diabetes who showed no significant differences in PPP and plantar tissue thickness compared to those without diabetes, several texture features of ultrasound images were found to be significantly different. Our finding indicates that texture features (correlation, energy and entropy) of macrochambers could be used for early detection of soft tissue damage associated with diabetes.

A Radiomics Model for the Differentiation of Intracranial Solitary Fibrous Tumor/Hemangiopericytoma and Meningioma Based on Multiparametric Magnetic Resonance Imaging.

Although the imaging findings of intracranial solitary fibrous tumor (SFT)/hemangiopericytoma (HPC) and meningioma are similar, their treatment and prognosis are quite different. Accurate preoperative identification of these two types of tumors is crucial for individualized treatment. The aim of this study was to develop a radiomics model for the differentiation of intracranial SFT/HPC and meningioma based on multiparametric magnetic resonance imaging (mpMRI). A total of 99 patients from July 2012 to July 2018 with histologically and immunohistochemically confirmed SFT/HPC (n = 40) or meningiomas (n = 59) were retrospectively analyzed. A total of 1118 features were extracted based on its image shape, intensity and texture features. The logistic regression (LR) and multi-layer artificial neural network (ANN) classifiers were used to classify SFT/HPC and meningioma. The predictive performance was calculated using receiver operating characteristic curves (ROC). We found significant difference in terms of sex between the SFT/HPC and meningioma group (χ2= 4.829, P < 0.05), but no significant difference was found in age (P > 0.05). The most significant radiomics features included five shape and four first-order level features. For the LR classifier, the prediction accuracy of SFT/HPC was 71.0% and meningioma was 78.7%. For the ANN classifier, the prediction accuracy of SFT/HPC was 83.9% and meningioma was 80.9%. Both of the two classifiers achieved a high accuracy rate, but ANN was better. Radiomics features, especially when combined with an ANN classifier, can provide satisfactory performance in distinguishing SFT/HPC and meningioma.

A hyperdimensional framework: Unveiling the interplay of RBP and GSN within CNNs for ultra-precise brain tumor classification

This study presents the RBP-CNN model, a convolutional neural network specifically designed for the precise classification of brain tumors in medical imaging. Conventional methods often encounter difficulties in extracting image noise and texture features, which has led to the incorporation of regional binary patterns (RBP) and Gray Standard Normalization (GSN) preprocessing techniques in CNN. The research addresses fundamental inquiries regarding the impact of the model on accuracy, false classifications, and efficiency. The novelty of RBP-CNN lies in its distinctive approach to extracting texture features, which involves optimizing pixel values through GSN preprocessing and generating regional binary patterns based on integral images. The objective of this research is to bridge a critical gap by providing a more accurate and efficient model for classifying brain tumors. The key findings reveal the exceptional performance of RBP-CNN, achieving a classification accuracy of 96% with a reduced false classification ratio of 7% across a dataset of 3000 samples. Comparative analyses position RBP-CNN as superior to alternative models in terms of accuracy, false classification rates, and efficiency. The structural insights and hyperparameter values of the model, as well as its application to the FigShare dataset, demonstrate its robustness and scalability. RBP-CNN emerges as an innovative and effective solution, advancing the field of medical image categorization. The findings of this study contribute a novel methodology, paving the way for future exploration in hyperspectral image applications and positioning RBP-CNN as a potential state-of-the-art tool for medical image analysis.

Exploring the Differences in Tree Species Classification between Typical Forest Regions in Northern and Southern China

Focusing on the trend of continuously seeking high-precision tree species classification results in small areas from the perspectives of sensors and classification algorithms. This study aimed to explore the effects of data sources, classifiers, and seasons on classification accuracy in regions with significant environmental variation, examining patterns of tree species classification to enhance the transferability of classification. Considering two typical forest distribution regions in the north and south of China, this study utilized the revisitation cycle and open-source advantages of Sentinel-2 and Landsat-8. Leveraging the Google Earth Engine (GEE) platform, this study captured spectral features, vegetation indices, and texture features for single seasonal and seasonal combination images. With the assistance of Sentinel-1A and SRTM (Shuttle Radar Topography Mission) DEM (Digital Elevation Model), backscattering coefficient features and topographical features were extracted and input with features captured from Sentinel-2 and Landsat-8 into three types of classifiers: random forest (RF), support vector machine (SVM), and gradient tree boosting (GTB) for major tree species classification. In this research, we discovered that the best classification for single season in the northern study area was spring, whereas, for the southern study area, it was winter. Seasonal combination images effectively improved the classification accuracy of single seasonal images, with Sentinel-2 imagery displaying better classification performance compared to Landsat-8, and the optimal classifier differing between the north and the south. The inclusion of topographical or backscattering coefficient features in the four-season combination imagery contributed to improvements in classification accuracy, with topographical features significantly enhancing the classification performance in the topographically varied southern study area. The evaluation of feature importance indicated that elevation was the most critical feature for classification, while spectral features and vegetation indices were also significant. In the southern study area with large topographical discrepancies, subdividing into different terrain units led to improved tree species classification accuracy in medium-altitude, gentle slope areas. These findings provide insights into the regularity of enhancing tree species classification accuracy in environmentally diverse areas through the use of multi-source remote sensing data and multi-seasonal imagery. Consequently, the results offer a reference for the identification of tree species across large areas and the creation of spatial distribution maps.

Detection of jelly orange granulation disease using a dual-input Resnet-Transformer model (DresT) based on acoustic vibration images and a novel acoustic vibration device

Granulation is a common internal disease in citrus fruits, and it is difficult to distinguish fruits with granulation disease from their appearance. In this study, a novel acoustic vibration device based on a micro-LDV, a microphone and a resonance speaker was employed to collect acoustic vibration response signals of "Aiyuan 38" jelly orange. The one-dimensional acoustic vibration response signal was converted into acoustic vibration images, and a double-input Resnet-Transformer network (DresT) was constructed for extracting deep features in acoustic vibration images for identifying jelly-orange granulation disease. Firstly, train Drest and Resnet50 models using acoustic vibration images and compare the performance of Drest with that of Resnet50 (based on CNN). Then PLS-DA and SVM models are trained using acoustic vibration image texture features or acoustic vibration spectral features, and the performance is compared with the DresT model. The results showed that the DresT model trained using acoustic vibration images can accurately identify jelly orange granulation disease with a detection accuracy of 99.31 %. The F1 of the model is 99.5 %, the accuracy is 99.01 %, and the recall is 100 %.