Explore the vast range of titles available.

Monitoring the growth conditions and behavior of fish will enable scientific management, reduce the threat of losses caused by disease and stress. Traditional monitoring methods are time-consuming, laborious, and untimely monitoring readily leads to aquaculture accidents. As a non-invasive, objective, and repeatable tool, machine vision systems have been widely used in various aspects of aquaculture monitoring. Nevertheless, the complex underwater environment makes it difficult to obtain ideal data processing results only using traditional image processing methods. Due to their powerful feature extraction capabilities, deep learning (DL) algorithms have been widely used in underwater image processing. Hence, the combination of DL algorithms and machine vision for the automated monitoring of aquaculture is of great importance. As evidence for the multidisciplinary aspects of DL applications, attention is focused on the latest DL methods applied to five fields of research: classification, detection, counting, behavior recognition, and biomass estimation. Meanwhile, due to the low training efficiency of DL models caused by insufficient dataset, transfer learning and GAN have also put into spotlight of this filed to pursue high performance of DL models. We also present the challenges and benchmarks in terms of the advantages and disadvantages of the selected method in each field. In addition, we review the sources of image acquisition and pre-processing methods in aquaculture. Finally, the challenges and prospects of DL in aquaculture machine vision systems are discussed. The literature review shows that the deep neural networks such as AlexNet, LSTM, VGG, and GoogLeNet, have been used for aquaculture machine vision systems.

To address the issue of low recognition accuracy of yarn tube types in practical industrial scenarios, this study proposes a SimAM-ResNet18-based image recognition method for cone yarns. Different from the traditional yarn recognition method based on Resnet, the framework introduced in this study combines parameter free attention and swish activation to improve the recognition accuracy and robustness under industrial conditions. First, a high-resolution image acquisition system was designed and implemented. The acquired images were preprocessed using bilateral filtering, Gamma correction, HSI color space extraction, and rapid template matching of edge points to enhance image features. Then, the Swish activation function and SimAM attention mechanism were integrated into the ResNet18 network, effectively improving the model's focus on key regions and its feature representation capabilities. On a dataset composed of 1800 real-world images collected from a textile production line, the proposed model achieved a recognition accuracy of 98.3%, a precision of 0.969, a recall of 0.972, and an F1-score of 0.970, significantly outperforming mainstream models such as MobileNetV2, EfficientNet-B0, and SENet18. Without retraining, the model maintained an accuracy of 92.8% under challenging conditions such as angle variation and illumination changes, demonstrating strong generalization capability and practical industrial value.

In hyperspectral transmission imaging (mainly refers to transmission breast imaging), the strong scattering characteristics of the tissue cause the blurred image and weak image signal, which hinders heterogeneity detection in tissue. In this paper, we designed the simulation experiment of collecting phantom images, and a joint preprocessing algorithm suitable for transmission tissue image is proposed and verified: the algorithm combining single channel frame accumulation and edge enhancement algorithm. The result shows that the PSNR of the phantom image is increased to 57.3 dB and the edge of phantom image processed by the joint preprocessing algorithm is preserved; the standard deviation is 19.8998 higher than original image, that is, the contrast is greatly improved. In our previous work, the detection accuracy of the image processed by this algorithm is higher than that without processed when the image detected in object detection algorithm based on deep learning; the mAP reaches 99.9%. Therefore, the preprocessing algorithm in this paper provides a highly compatible and easier preprocessing method for heterogeneity detection of multispectral tissue images, which improves the detection accuracy of heterogeneity to some extent. And it may be a new way to improve the quality of such multispectral and hyperspectral transmission tissue images.

The foundation of effectively predicting plant disease in the early stage using deep learning algorithms is ideal for addressing food insecurity, inevitably drawing researchers and agricultural specialists to contribute to its effectiveness. The input preprocessor, abnormalities of the data (i.e., incomplete and nonexistent features, class imbalance), classifier, and decision explanation are typical components of a plant disease detection pipeline based on deep learning that accepts an image as input and outputs a diagnosis. Data sets related to plant diseases frequently display a magnitude imbalance due to the scarcity of disease outbreaks in real field conditions. This study examines the effects of several preprocessing methods and class imbalance approaches and deep learning classifiers steps in the pipeline for detecting plant diseases on our data set. We notably want to evaluate if additional preprocessing and effective handling of data inconsistencies in the plant disease pipeline may considerably assist deep learning classifiers. The evaluation’s findings indicate that contrast limited adaptive histogram equalization (CLAHE) combined with image sharpening and generative adversarial networks (GANs)-based approach for resampling performed the best among the preprocessing and resampling techniques, with an average classification accuracy of 97.69% and an average F1-score of 97.62% when fed through a ResNet-50 as the deep learning classifier. Lastly, this study provides a general workflow of a disease detection system that allows each component to be individually focused on depending on necessity.

Automatic text recognition from the natural images acquired in uncontrolled lighting conditions is a challenging task due to the presence of shadows hindering the shape analysis and classification of individual characters. Since the optical character recognition methods require prior image binarization, the application of classical global thresholding methods in such case makes it impossible to preserve the visibility of all characters. Nevertheless, the use of adaptive binarization does not always lead to satisfactory results for heavily unevenly illuminated document images. In this paper, the image preprocessing methodology with the use of local image entropy filtering is proposed, allowing for the improvement of various commonly used image thresholding methods, which can be useful also for text recognition purposes. The proposed approach was verified using a dataset of 140 differently illuminated document images subjected to further text recognition. Experimental results, expressed as Levenshtein distances and F-Measure values for obtained text strings, are promising and confirm the usefulness of the proposed approach.

The increasing incidence of skin cancer necessitates advancements in early detection methods, where deep learning can be beneficial. This study introduces SkinNet-14, a novel deep learning model designed to classify skin cancer types using low-resolution dermoscopy images. Unlike existing models that require high-resolution images and extensive training times, SkinNet-14 leverages a modified compact convolutional transformer (CCT) architecture to effectively process 32 × 32 pixel images, significantly reducing the computational load and training duration. The framework employs several image preprocessing and augmentation strategies to enhance input image quality and balance the dataset to address class imbalances in medical datasets. The model was tested on three distinct datasets—HAM10000, ISIC and PAD—demonstrating high performance with accuracies of 97.85%, 96.00% and 98.14%, respectively, while significantly reducing the training time to 2–8 s per epoch. Compared to traditional transfer learning models, SkinNet-14 not only improves accuracy but also ensures stability even with smaller training sets. This research addresses a critical gap in automated skin cancer detection, specifically in contexts with limited resources, and highlights the capabilities of transformer-based models that are efficient in medical image analysis.

Brain tumors pose a significant medical challenge necessitating precise detection and diagnosis, especially in Magnetic resonance imaging(MRI). Current methodologies reliant on traditional image processing and conventional machine learning encounter hurdles in accurately discerning tumor regions within intricate MRI scans, often susceptible to noise and varying image quality. The advent of artificial intelligence (AI) has revolutionized various aspects of healthcare, providing innovative solutions for diagnostics and treatment strategies. This paper introduces a novel AI-driven methodology for brain tumor detection from MRI images, leveraging the EfficientNetB2 deep learning architecture. Our approach incorporates advanced image preprocessing techniques, including image cropping, equalization, and the application of homomorphic filters, to enhance the quality of MRI data for more accurate tumor detection. The proposed model exhibits substantial performance enhancement by demonstrating validation accuracies of 99.83%, 99.75%, and 99.2% on BD-BrainTumor, Brain-tumor-detection, and Brain-MRI-images-for-brain-tumor-detection datasets respectively, this research holds promise for refined clinical diagnostics and patient care, fostering more accurate and reliable brain tumor identification from MRI images. All data is available on Github: https://github.com/muskan258/Brain-Tumor-Detection-from-MRI-Images-Utilizing-EfficientNetB2 ).

Carbon fiber-reinforced polymer (CFRP) laminates are widely used in aerospace, automotive, and infrastructure industries due to their high strength-to-weight ratio. However, defect detection in CFRP remains challenging, particularly in low signal-to-noise ratio (SNR) conditions. Conventional segmentation methods often struggle with noise interference and signal variations, leading to reduced detection accuracy. In this study, we evaluate the impact of thermal image preprocessing on improving defect segmentation in CFRP laminates inspected via pulsed thermography. Polynomial approximations and first- and second-order derivatives were applied to refine thermographic signals, enhancing defect visibility and SNR. The U-Net architecture was used to assess segmentation performance on datasets with and without preprocessing. The results demonstrated that preprocessing significantly improved defect detection, achieving an Intersection over Union (IoU) of 95% and an F1-Score of 99%, outperforming approaches without preprocessing. These findings emphasize the importance of preprocessing in enhancing segmentation accuracy and reliability, highlighting its potential for advancing non-destructive testing techniques across various industries.

Skin cancer stands as one of the foremost challenges in oncology, with its early detection being crucial for successful treatment outcomes. Traditional diagnostic methods depend on dermatologist expertise, creating a need for more reliable, automated tools. This study explores deep learning, particularly Convolutional Neural Networks (CNNs), to enhance the accuracy and efficiency of skin cancer diagnosis. Leveraging the HAM10000 dataset, a comprehensive collection of dermatoscopic images encompassing a diverse range of skin lesions, this study introduces a sophisticated CNN model tailored for the nuanced task of skin lesion classification. The model’s architecture is intricately designed with multiple convolutional, pooling, and dense layers, aimed at capturing the complex visual features of skin lesions. To address the challenge of class imbalance within the dataset, an innovative data augmentation strategy is employed, ensuring a balanced representation of each lesion category during training. Furthermore, this study introduces a CNN model with optimized layer configuration and data augmentation, significantly boosting diagnostic precision in skin cancer detection. The model’s learning process is optimized using the Adam optimizer, with parameters fine-tuned over 50 epochs and a batch size of 128 to enhance the model’s ability to discern subtle patterns in the image data. A Model Checkpoint callback ensures the preservation of the best model iteration for future use. The proposed model demonstrates an accuracy of 97.78% with a notable precision of 97.9%, recall of 97.9%, and an F2 score of 97.8%, underscoring its potential as a robust tool in the early detection and classification of skin cancer, thereby supporting clinical decision-making and contributing to improved patient outcomes in dermatology.