Image Size Research Articles

A new 2D hyperchaotic model-based encryption strategy for multiple images via pixel fusion

Abstract In view of the intrinsic characteristics of chaotic system, it can be widely applied in image encryption. Considering the common issues of insufficient chaotic performance and discontinuous range of control parameters in current chaotic systems, a hybrid chaotic model with the characteristics of simple and easy to design is devised to address these challenges. Through phase portraits, bifurcation diagrams, and the analysis of Lyapunov exponent spectrum, it is demonstrated that the proposed system can exhibit robust traversal performance. To meet the demands of digital image encryption, an effective encryption strategy for multi-image based on this model is proposed, which can accommodate the varying image sizes and types simultaneously. In the given strategy, the pixels are merged from multiple images at first, and then the keys can be derived with SHA-512. By the application of discrete wavelet transform, the information can be embedded from the plaintext images, which can ensure the robust resistance to data loss while facilitating the shared transmission of plaintext information. Finally, by leveraging the developed chaotic system, a three-dimensional cross-plane scrambling and diffusion algorithm is introduced to enhance the effectiveness of encryption scheme.

Protocol for quantifying muscle fiber size, number, and central nucleation of mouse skeletal muscle cross-sections using Myotally software.

Here, we present a protocol for using Myotally, a user-friendly software for fast, automated quantification of muscle fiber size, number, and central nucleation from immunofluorescent stains of mouse skeletal muscle cross-sections. We describe steps for installing the software, preparing compatible images, finding the file path, and selecting key parameters like image quality and size limits. We also detail optional features, such as measuring mean fluorescence. By automating these traditionally labor-intensive processes, Myotally improves research efficiency and data consistency.

Enhancing kelp forest detection in remote sensing images using crowdsourced labels with Mixed Vision Transformers and ConvNeXt segmentation models

ABSTRACT Kelp forests, as foundation species, are vital to marine ecosystems, providing essential food and habitat for numerous organisms. This study explores the integration of crowdsourced labels with advanced artificial intelligence models to develop a fast and accurate kelp canopy detection pipeline using Landsat imagery. Building on the success of a machine learning competition, where this approach ranked third and performed consistently well on both local validation and public and private leaderboards, the research highlights the effectiveness of combining Mixed Vision Transformers (MIT) with ConvNeXt models. Training these models on various image sizes significantly enhanced the accuracy of the ensemble results. U-Net emerged as the best segmentation architecture, with UpperNet also contributing to the final ensemble. Key Landsat bands, such as ShortWave InfraRed (SWIR1) and Near-InfraRed (NIR), were crucial while altitude data were used in postprocessing to eliminate false positives on land. The methodology achieved a high detection rate, accurately identifying about three out of four pixels containing kelp canopy while keeping false positives low. Despite the medium resolution of Landsat satellites, their extensive historical coverage makes them effective for studying kelp forests. This work also underscores the potential of combining machine learning models with crowdsourced data for effective and scalable environmental monitoring. All running code for training all models and inference can be found at https://github.com/IoannisNasios/Kelp_Forests.

ENSEMBLE METHOD BASED ON AVERAGING SHAPES OF OBJECTS USING THE PYRAMID METHOD

Context. Image segmentation plays a key role in computer vision. The quality of segmentation is affected by many factors: noise, artifacts, complex shapes of objects. Classical methods cannot always guarantee good success, depending on the quality of the image and the existing noise, they cannot always achieve the desired result. The proposed method uses an ensemble of neural networks, which makes it possible to increase the accuracy and stability of segmentation. Objective. The goal of the work is to develop a new method of combining predictions of neural network ensembles, which can improve segmentation accuracy by combining images of different image sizes. Method. A method is proposed that averages the shapes of objects depicted on prediction masks. A pyramid of images is used to improve segmentation quality, each level of the pyramid corresponds to an increased size of the original image. This approach allows obtaining image characteristics at different levels. For a test image, a prediction is obtained from each neural network in the ensemble, after which a pyramid is built for the image. All pyramid levels are combined into the final image using SAAMC. All obtained final images for each neural network are also combined at the end using SAAMC. The use of an ensemble of neural networks combined with the pyramid method allows for reducing the impact of noise and artifacts on the segmentation results. Results. The use of this method was compared with the usual use of individual neural networks and the ensemble averaging method. The obtained results show that the proposed method outperforms its competitors. Application of the proposed method improved the accuracy and quality of segmentation. Conclusions. The conducted research confirmed the sense of using an ensemble of neural networks and creating a new method of combining predictions. The use of an ensemble of neural networks makes it possible to compensate for the errors and shortcomings of individual neural networks. Using the proposed method can significantly reduce the impact of noise and artifacts on segmentation. Further study and modification of this method will make it possible to further improve the quality of segmentation.

Evaluasi Strategi Bersaing Teh Botol Sosro di Kota Surakarta (Pendekatan Competitive Profile Matrix)

The increasing popularity of packaged tea beverages has encouraged many companies to launch tea drink products and compete to gain positive consumer responses. This competitive condition demands that Teh Botol Sosro understand its market position and develop effective strategies to maintain and expand its market share. Therefore, this study applies the Competitive Profile Matrix (CPM) to analyze Teh Botol Sosro's market position in Surakarta, the first study in this context. The research aims to identify the key competitive factors, evaluate Teh Botol Sosro's position, and formulate alternative strategies to enhance its competitiveness. The research method is descriptive, with a purposive location selection in Surakarta. Key informants include the Area Sales Manager, Sales Supervisor, and Distributor Supervisor of PT Sinar Sosro in Surakarta City, with samples selected using accidental sampling. The data used includes primary and secondary data collected through observation, interviews, and documentation, then analyzed using the CPM analytical tool. Research findings showed that the primary competitive factors for Teh Botol Sosro include freshness of taste, tea fragrance, product size variations, packaging design, price, advertising, brand image, discounts, and distribution. Based on the CPM, Teh Botol Sosro ranks second, excelling in brand image, advertising, packaging design, and product size variations, but still competes closely with Teh Pucuk Harum, which holds the first position. The recommended alternative strategies include (1) maintaining performance on superior attributes, (2) benchmarking against competitors in aspects of freshness of taste, tea fragrance, nutrition, and price, and (3) enhancing discount and distribution strategies. This study is expected to provide insights into understanding the factors influencing consumer preferences and support the development of adaptive agribusiness strategies in the ready-to-drink tea industry.

The Measurement of Metal Mineral Particle Size Under the Microscope Based on Gaussian Pyramids and Directional Maximum Intercept

With the development of mineral resources, minerals are becoming increasingly difficult to process. In order to utilize these resources more effectively, in-depth research into process mineralogy has become increasingly important in the field of mineralogy, and particle size measurement under the microscope is one of the critical aspects of process mineralogy. At present, the use of scanning electron microscopes and other equipment for measurement is very expensive, and manual measurement has problems such as poor accuracy and low efficiency. In addition, there is a lack of reference materials for the segmentation algorithm of mineral light images. This article proposes a Gaussian pyramid based on bilateral filtering combined with directional maximum intercept to measure mineral particle size under the microscope. In the experiments, different segmentation algorithms were studied, including Gaussian pyramid segmentation based on bilateral filtering, segmentation based on Fuzzy C-Means, and the rapidly developing deep learning segmentation algorithms in recent years. By comparing the segmentation effects of these three algorithms on various mineral thin-section images, the Gaussian pyramid segmentation algorithm based on bilateral filtering was selected as the optimal one. This was then combined with the directional maximum intercept method to measure the particle size of ilmenite and pyrite images. The experimental results show that the segmentation method based on the bilateral filtering Gaussian pyramid technique has higher segmentation accuracy than the other two algorithms, and can accurately measure the particle size of minerals under the microscope. Compared with manual measurement, this method can effectively and accurately measure the microscopic particle size of target minerals, greatly reducing the workload of measurement personnel and reducing the time spent on measurement.

Tuberculosis Detection based on Lung X-Ray Images Using Convolutional Neural Networks (CNN)

Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis, primarily affecting the lungs. Despite being preventable and curable, TB remains a significant global health issue, especially in developing countries. The success of TB treatment heavily depends on the accuracy of the diagnosis, which typically requires expertise from pulmonology or radiology specialists to interpret chest X-ray images. This study aims to design an assistive tool for TB detection that can automatically diagnose the disease using chest X-ray data. The study implemented a Convolutional Neural Network (CNN) architecture to analyze the X-ray images. Additionally, image preprocessing and early stopping methods were employed to enhance accuracy performance, optimize computation, and prevent overfitting. Experiment was conducting using 75% of the data as training data to generate the model and then applied to 25% of the data as testing data. This study comparing image sizes in RGB and grayscale modes. Experimental results show that the use of early stopping has a significant impact on training time, reducing training time substantially in almost all scenarios without drastically sacrificing accuracy. Without early stopping, accuracy does tend to be higher, as seen in grayscale color mode with an image size of 128x128, where the accuracy reaches 0.992, and in RGB mode with an image size of 64x64 which reaches 0.995. However, training time also increases significantly, for example for a 299x299 image with RGB mode, the training time reaches 927 seconds. Therefore, while RGB yields slightly higher accuracy, grayscale is recommended due to significantly faster training times. Additionally, the early stopping mechanism proves effective in reducing computational time, making the training process more efficient.

Deep Learning for Green Growth: Custom CNN Uncovers Cucumber Leaf Disease Secrets

There are many popular vegetables in the world, among them cucumber is one of the most famous and healthiest foods. It comes with multiple health benefits, especially in losing weight and in skincare. However, there can be multiple diseases that can attack the cucumber leaf and the cucumber itself. Cucumber leaf disease research plays a cardinal role in agricultural fields. Cucumber leaf disease is a special disease that harms its leaves and causes different types of diseases. In real fields, it is difficult to identify different diseases on the leaf due to the complex environment. It is also a very challenging and time-consuming process to identify leaf diseases. To overcome those barriers and detect the diseases we have used two learning methods namely DenseNet201 and VGG16 model. In this research, we have created a new dataset with four classes containing 2159 pictures with an accuracy rate of 98.97% from DenseNet201. The image recognition methods choose the image and find the disease's spot by breaking the image frame by frame. We have rescaled all the images while preprocessing where the image size is 224×224. The methods show great success in finding the disease spots.

Magnetic resonance image denoising for Rician noise using a novel hybrid transformer-CNN network (HTC-net) and self-supervised pretraining.

Magnetic resonance imaging (MRI) is a crucial technique for both scientific research and clinical diagnosis. However, noise generated during MR data acquisition degrades image quality, particularly in hyperpolarized (HP) gas MRI. While deep learning (DL) methods have shown promise for MR image denoising, most of them fail to adequately utilize the long-range information which is important to improve denoising performance. Furthermore, the sample size of paired noisy and noise-free MR images also limits denoising performance. To develop an effective DL method that enhances denoising performance and reduces the requirement of paired MR images by utilizing the long-range information and pretraining. In this work, a hybrid Transformer-convolutional neural network (CNN) network (HTC-net) and a self-supervised pretraining strategy are proposed, which effectively enhance the denoising performance. In HTC-net, a CNN branch is exploited to extract the local features. Then a Transformer-CNN branch with two parallel encoders is designed to capture the long-range information. Within this branch, a residual fusion block (RFB) with a residual feature processing module and a feature fusion module is proposed to aggregate features at different resolutions extracted by two parallel encoders. After that, HTC-net exploits the comprehensive features from the CNN branch and the Transformer-CNN branch to accurately predict noise-free MR images through a reconstruction module. To further enhance the performance on limited MRI datasets, a self-supervised pretraining strategy is proposed. This strategy employs self-supervised denoising to equip the HTC-net with denoising capabilities during pretraining, and then the pre-trained parameters are transferred to facilitate subsequent supervised training. Experimental results on the pulmonary HP 129Xe MRI dataset (1059 images) and IXI dataset (5000 images) all demonstrate the proposed method outperforms the state-of-the-art methods, exhibiting superior preservation of edges and structures. Quantitatively, on the pulmonary HP 129Xe MRI dataset, the proposed method outperforms the state-of-the-art methods by 0.254-0.597dB in PSNR and 0.007-0.013 in SSIM. On the IXI dataset, the proposed method outperforms the state-of-the-art methods by 0.3-0.927dB in PSNR and 0.003-0.016 in SSIM. The proposed method can effectively enhance the quality of MR images, which helps improve the diagnosis accuracy in clinical.

Low-Quality Sensor Data-Based Semi-Supervised Learning for Medical Image Segmentation.

Traditional medical image sensors face multiple challenges. First, these sensors typically rely on large amounts of labeled data, which are time-consuming and costly to obtain. Second, when the data volume and image size are large, traditional sensors have limited computational power, making it difficult to effectively train and infer models. Additionally, traditional sensors have poor generalization ability and struggle to adapt to datasets with different modalities. This paper devises a novel framework, named LSDSL, and deploys it in the sensor. LSDSL utilizes low-quality sensor data for semi-supervised learning in medical image segmentation. in supervised learning, we devise the hard region exploration (hre) module to enhance the model's comprehension of low-quality pixels in hard regions. in unsupervised learning, we introduce a pseudo-label sharing (ps) module, which allows low-quality pixels in one network to learn from the high-quality pixels in the other networks. our model outperforms other semi-supervised methods on the datasets of two different modalities (CT and MRI) in medical image sensors, achieving superior inference speed and segmentation accuracy.

An Evaluation of Image Slicing and YOLO Architectures for Object Detection in UAV Images

Object detection in aerial images poses significant challenges due to the high dimensions of the images, requiring efficient handling and resizing to fit object detection models. The image-slicing approach for object detection in aerial images can increase detection accuracy by eliminating pixel loss in high-resolution image data. However, determining the proper dimensions to slice is essential for the integrity of the objects and their learning by the model. This study presents an evaluation of the image-slicing approach for alternative sizes of images to optimize efficiency. For this purpose, a dataset of high-resolution images collected with Unmanned Aerial Vehicles (UAV) has been used. The experiments evaluated using alternative YOLO architectures like YOLOv7, YOLOv8, and YOLOv9 show that the image dimensions significantly change the performance results. According to the experiments, the best mAP@05 accuracy was obtained by slicing 1280×1280 for YOLOv7 producing 88.2. Results show that edge-related objects are better preserved as the overlap and slicing sizes increase, resulting in improved model performance.

Development of an infrared diagnostic system with multi-spectrum or high frame rate for measuring the synchrotron radiation emitted by runaway electrons on EAST

In order to meet the requirements for high frame rate and synchrotron radiation spectrum of infrared diagnostic systems when studying runaway electron, an infrared diagnostic system with multi-spectrum or high frame rate has been developed and installed on the Experimental Advanced Superconducting Tokamak(EAST) for studying the runaway electrons. It is based on a high frame rate infrared camera with a high-speed rotating filter wheel and an endoscope. The infrared camera can achieve a maximum frame rate of 120 kHz in sub-windowing mode with an image size of 64 × 2. In experiments, effective data has been obtained at a frame rate of 10 kHz and an image size of 64 × 4. This diagnostic system may become a valuable tool for studying the interaction between runaway electrons and magnetohydrodynamic instability in the future. The equipped high-speed rotating filter wheel on the camera can be utilized to capture the synchrotron radiation spectrum of runaway electrons, which has been calibrated using a blackbody furnace. The synchrotron radiation spectrum contains valuable information about runaway electrons and can be one of the important parameters for inferring their accurate distribution function in the future.

DAC: Differentiable Auto-Cropping in Deep Learning

Auto-cropping, the process of automatically adjusting the boundaries of an image to focus on the region of interest, is crucial to improving the diagnostic quality of dental panoramic radiographs. Its importance lies in its ability to standardize the size of different input images with minimal loss of information, thus ensuring consistency and improving the performance of subsequent image-processing tasks. Despite the widespread use of CNNs in many studies, research on auto-cropping for different-sized images remains limited. This study aims to explore the potential of differentiable auto-cropping in dental panoramic radiographs. A unique dataset of 20,973 dental panoramic radiographs, mostly with a resolution of 2836×1536 or close, divided into five classes by 3 dentists, was used, which is the same dataset from the previous study (Top et al. 2023). ResNet-101 model, which was the most successful network for the dataset (Top et al. 2023), was used for the evaluation. To reduce variance, the model was evaluated using 10-fold cross-validation for both non-auto-cropped and auto-cropped trainings. Data augmentation was also used to produce more accurate and robust results. For auto-cropped training, it was adjusted to be much less effective than the non-auto-cropped one. Accuracy was improved by 1.8%, from 92.7% to 94.5%, thanks to the proposed auto-crop optimization developed to reduce dataset-related issues. Its macro-average AUC was also raised from 0.989 to 0.993. The proposed auto-crop optimization can be implemented as a trainable network layer in an end-to-end CNN and can be used for other problems as well. Increasing the accuracy from 92.7% to 94.5% is a very challenging task due to diminishing returns, as there is little room for improvement. The results show the potential of the proposed differentiable auto-crop algorithm and encourages its use in different fields.

Contextual Learning in Fourier Complex Field for VHR Remote Sensing Images.

Very high-resolution (VHR) remote sensing (RS) image classification is the fundamental task for RS image analysis and understanding. Recently, Transformer-based models demonstrated outstanding potential for learning high-order contextual relationships from natural images with general resolution ( ≈ 224 × 224 pixels) and achieved remarkable results on general image classification tasks. However, the complexity of the naive Transformer grows quadratically with the increase in image size, which prevents Transformer-based models from VHR RS image ( ≥ 500 × 500 pixels) classification and other computationally expensive downstream tasks. To this end, we propose to decompose the expensive self-attention (SA) into real and imaginary parts via discrete Fourier transform (DFT) and, therefore, propose an efficient complex SA (CSA) mechanism. Benefiting from the conjugated symmetric property of DFT, CSA is capable to model the high-order contextual information with less than half computations of naive SA. To overcome the gradient explosion in Fourier complex field, we replace the Softmax function with the carefully designed Logmax function to normalize the attention map of CSA and stabilize the gradient propagation. By stacking various layers of CSA blocks, we propose the Fourier complex Transformer (FCT) model to learn global contextual information from VHR aerial images following the hierarchical manners. Universal experiments conducted on commonly used RS classification datasets demonstrate the effectiveness and efficiency of FCT, especially on VHR RS images. The source code of FCT will be available at https://github.com/Gao-xiyuan/FCT.

Pixel Distillation: Cost-Flexible Distillation Across Image Sizes and Heterogeneous Networks.

Previous knowledge distillation (KD) methods mostly focus on compressing network architectures, which is not thorough enough in deployment as some costs like transmission bandwidth and imaging equipment are related to the image size. Therefore, we propose Pixel Distillation that extends knowledge distillation into the input level while simultaneously breaking architecture constraints. Such a scheme can achieve flexible cost control for deployment, as it allows the system to adjust both network architecture and image quality according to the overall requirement of resources. Specifically, we first propose an input spatial representation distillation (ISRD) mechanism to transfer spatial knowledge from large images to student's input module, which can facilitate stable knowledge transfer between CNN and ViT. Then, a Teacher-Assistant-Student (TAS) framework is further established to disentangle pixel distillation into the model compression stage and input compression stage, which significantly reduces the overall complexity of pixel distillation and the difficulty of distilling intermediate knowledge. Finally, we adapt pixel distillation to object detection via an aligned feature for preservation (AFP) strategy for TAS, which aligns output dimensions of detectors at each stage by manipulating features and anchors of the assistant. Comprehensive experiments on image classification and object detection demonstrate the effectiveness of our method.

Development of Software Interface for AI-Driven Weed Control in Robotic Vehicles, with Time-Based Evaluation in Indoor and Field Settings

Herbicide blanket application is widely practiced by farmers to chemically control weeds in the field, thereby enhancing productivity and improving crop quality. However, their repetitive use also has caused several negative issues on the environment and human health. To overcome these negative issues precision site-specific weed control can be used to reduce the herbicide application by targeted application on only weed areas. Hence, a robotic platform utilizing NVIDIA Jetson embedded device as control and processing unit was developed and tested on lab and field conditions for weed control. The average time taken for data acquisition and artificial intelligence-based computer vision tasks was tested under both lab and field conditions. Moreover, a grid map creation algorithm using YOLOv4 deep learning algorithm was evaluated to control the nozzles of the robotic platform. Integrating all these components together enabled real-time weed management approaches, enhancing the precision and efficiency of herbicide application Independent paired t-tests reveal that there is no significant difference in computational time between lab and field testing. Conversely, independent paired t-test revels that there is significant difference in image size between lab and field testing. The reduction of herbicide application based on grid map was obtained 79 to 80 %, which does not include YOLOv4 algorithm failure and system synchronization failure. The study suggests the importance of field testing for real-time applications of the robotic platform by using deep learning computer vision methods for weed control.

A Parkinson's disease-related nuclei segmentation network based on CNN-Transformer interleaved encoder with feature fusion.

Automatic segmentation of Parkinson's disease (PD) related deep gray matter (DGM) nuclei based on brain magnetic resonance imaging (MRI) is significant in assisting the diagnosis of PD. However, due to the degenerative-induced changes in appearance, low tissue contrast, and tiny DGM nuclei size in elders' brain MRI images, many existing segmentation models are limited in the application. To address these challenges, this paper proposes a PD-related DGM nuclei segmentation network to provide precise prior knowledge for aiding diagnosis PD. The encoder of network is designed as an alternating encoding structure where the convolutional neural network (CNN) captures spatial and depth texture features, while the Transformer complements global position information between DGM nuclei. Moreover, we propose a cascaded channel-spatial-wise block to fuse features extracted by the CNN and Transformer, thereby achieving more precise DGM nuclei segmentation. The decoder incorporates a symmetrical boundary attention module, leveraging the symmetrical structures of bilateral nuclei regions by constructing signed distance maps for symmetric differences, which optimizes segmentation boundaries. Furthermore, we employ a dynamic adaptive region of interests weighted Dice loss to enhance sensitivity towards smaller structures, thereby improving segmentation accuracy. In qualitative analysis, our method achieved optimal average values for PD-related DGM nuclei (DSC: 0.854, IOU: 0.750, HD95: 1.691 mm, ASD: 0.195 mm). Experiments conducted on multi-center clinical datasets and public datasets demonstrate the good generalizability of the proposed method. Furthermore, a volumetric analysis of segmentation results reveals significant differences between HCs and PDs. Our method holds promise for assisting clinicians in the rapid and accurate diagnosis of PD, offering a practical method for the imaging analysis of neurodegenerative diseases.

Accurate Lung Nodule Segmentation With Detailed Representation Transfer and Soft Mask Supervision.

Accurate lung lesion segmentation from computed tomography (CT) images is crucial to the analysis and diagnosis of lung diseases, such as COVID-19 and lung cancer. However, the smallness and variety of lung nodules and the lack of high-quality labeling make the accurate lung nodule segmentation difficult. To address these issues, we first introduce a novel segmentation mask named "soft mask", which has richer and more accurate edge details description and better visualization, and develop a universal automatic soft mask annotation pipeline to deal with different datasets correspondingly. Then, a novel network with detailed representation transfer and soft mask supervision (DSNet) is proposed to process the input low-resolution images of lung nodules into high-quality segmentation results. Our DSNet contains a special detailed representation transfer module (DRTM) for reconstructing the detailed representation to alleviate the small size of lung nodules images and an adversarial training framework with soft mask for further improving the accuracy of segmentation. Extensive experiments validate that our DSNet outperforms other state-of-the-art methods for accurate lung nodule segmentation, and has strong generalization ability in other accurate medical segmentation tasks with competitive results. Besides, we provide a new challenging lung nodules segmentation dataset for further studies (https://drive.google.com/file/d/15NNkvDTb_0Ku0IoPsNMHezJR TH1Oi1wm/view?usp=sharing).

Classification of COVID-19 Disease Based on Extra Tree Features Selection

The World Health Organization (WHO) has classified coronavirus as a global health emergency. Chest X-rays have been proven to be beneficial in both diagnosing and monitoring various lung diseases, including COVID-19. In this study, a COVID-19 disease detection framework is provided based on the methods used in machine learning and the Extra Tree algorithm to reduce the features extracted from the images. Using the Gray-Level Co-occurrence Matrix (GLCM) algorithm and the Extra Tree algorithm to choose the features of the work, a set of features is extracted from the images. This set of features is then put into the XGBoost algorithm to be classified. The proposed system was evaluated using two different sets of databases: the large database with 9544 images and the small database with 800 images. All image sizes were set to 300 x 300 pixels. The proposed system achieved a classification accuracy score of 90.04% using the large data set and 99.37% using the small set.

Ship detection using ensemble deep learning techniques from synthetic aperture radar imagery

Synthetic Aperture Radar (SAR) integrated with deep learning has been widely used in several military and civilian applications, such as border patrolling, to monitor and regulate the movement of people and goods across land, air, and maritime borders. Amongst these, maritime borders confront different threats and challenges. Therefore, SAR-based ship detection becomes essential for naval surveillance in marine traffic management, oil spill detection, illegal fishing, and maritime piracy. However, the model becomes insensitive to small ships due to the wide-scale variance and uneven distribution of ship sizes in SAR images. This increases the difficulties associated with ship recognition, which triggers several false alarms. To effectively address these difficulties, the present work proposes an ensemble model (eYOLO) based on YOLOv4 and YOLOv5. The model utilizes a weighted box fusion technique to fuse the outputs of YOLOv4 and YOLOv5. Also, a generalized intersection over union loss has been adopted in eYOLO which ensures the increased generalization capability of the model with reduced scale sensitivity. The model has been developed end-to-end, and its performance has been validated against other reported results using an open-source SAR-ship dataset. The obtained results authorize the effectiveness of eYOLO in multi-scale ship detection with an F1score and mAP of 91.49% and 92.00%, respectively. This highlights the efficacy of eYOLO in multi-scale ship detection using SAR imagery.