Pixel Classification Research Articles

Restoration and content analysis of ancient manuscripts via color space based segmentation.

Ancient manuscripts are a rich source of history and civilization. Unfortunately, these documents are often affected by different age and storage related degradation which impinge on their readability and information contents. In this paper, we propose a document restoration method that removes the unwanted interfering degradation patterns from color ancient manuscripts. We exploit different color spaces to highlight the spectral differences in various layers of information usually present in these documents. At each image pixel, the spectral representations of all color spaces are stacked to form a feature vector. PCA is applied to the whole data cube to eliminate correlation of the color planes and enhance separation among the patterns. The reduced data cube, along with the pixel spatial information, is used to perform a pixel based segmentation, where each cluster represents a class of pixels that share similar color properties in the decorrelated color spaces. The interfering, unwanted classes can thus be removed by inpainting their pixels with the background texture. Assuming Gaussian distributions for the various classes, a Gaussian Mixture Model (GMM) is estimated through the Expectation Maximization (EM) algorithm from the data, and then used to find appropriate labels for each pixel. In order to preserve the original appearance of the document and reproduce the background texture, the detected degraded pixels are replaced based on Gaussian conditional simulation, according to the surrounding context. Experiments are shown on manuscripts affected by different kinds of degradations, including manuscripts from the DIBCO 2018 and 2019 publicaly available dataset. We observe that the use of a few PCA dominant components accelerates the clustering process and provides a more accurate segmentation.

High-Precision 3D-DIC Measurement Method Based on Improved Forward Newton Iteration.

To solve the problems of the traditional 3D-DIC algorithm based on feature information or FFT search at the expense of accuracy in exchange for time, such as error-point extraction, mismatching of feature points, poor robustness, and accuracy loss caused by poor anti-noise performance, an improved high-precision 3D-DIC measurement method was proposed. In this method, the exact initial value is obtained by an exhaustive search. Then, the forward Newton iteration method is used for pixel classification, and the first-order nine-point interpolation is designed, which can quickly obtain the elements of Jacobian and Hazen matrix, and achieve accurate sub-pixel positioning. The experimental results show that the improved method has high accuracy, and its mean error and standard deviation stability and extreme value are better than similar algorithms. Compared with the traditional forward Newton method, the total iteration time of the improved forward Newton method is reduced in the subpixel iteration stage, and the computational efficiency is 3.8 times that of the traditional NR algorithm. The whole process of the proposed algorithm is simple and efficient, and it has application value in the precision occasions requiring high precision.

QE-DAL: A quantum image feature extraction with dense distribution-aware learning framework for object counting and localization

Object counting and localization (OCL) was an essential problem in intelligent transportation fields. The convolutional neural network (CNN)-based models transformed the OCL problems into a regression task. However, the abundant semantic information of the crowd scenes may lead the CNN framework hard to extract adequate features in order to ensure good precision In this work, a Quantum Image Feature Extraction with Dense Distribution-Aware Learning (QE-DAL) framework was proposed to handle this problem. The crowd features were extracted by Quantum layers, which were extracted by encoding, quantum circuits and decode procedures based on the multi-scale architecture. For handling objects, the refined distance compensating operator was adopted to fuse the multi-scale architecture. To relieve the computation burden, a Gaussian distribution estimation mechanism was proposed to initiate and update the bounding sizes of the objects via a point-supervised manner. Finally, the joint loss function, which describes pixel classes, density maps and offset bounding boxes, was built for QE-DAL. The ablation experiment results demonstrated that the effectiveness of the quantum feature extraction architecture and the Gaussian distribution estimation mechanism of QE-DAL was validated to show superior performance than the other state-of-the-art framework. Moreover, the generalization of the QE-DAL was evidenced by the Cross-scene learning evaluation.

IMAGE SEGMENTATION WITH A CONVOLUTIONAL NEURAL NETWORK WITHOUT POOLING LAYERS IN DERMATOLOGICAL DISEASE DIAGNOSTICS SYSTEMS

Context. The problem of automating of the segmentation of spectral-statistical texture images is considered. The object of research is image processing in dermatological disease diagnostic systems.
 Objective. The aim of the research is to improve the segmentation performance of color images of psoriasis lesions by elaboration of a deep learning convolutional neural network without pooling layers.
 Method. The convolutional neural network is proposed to process a three-channel psoriasis image with a specified size. The initial color images were scaled to the specified size and then inputed on the neural network. The architecture of the proposed neural network consists of four convolutional layers with batch normalization layers and ReLU activation function. Feature maps from the output of these layers were inputted to the 1*1 convolutional layer with the Softmax activation function. The resulting feature maps were inputted to the image pixel classification layer. When segmenting images, convolutional and pooling layers extract the features of image fragments, and fully connected layers classify the resulting feature vectors, forming a partition of the image into homogeneous segments. The segmentation features are evaluated as a result of network training using ground-truth images which segmented by an expert. Such features are robust to noise and distortion in images. The combination of segmentation results at different scales is determined by the network architecture. Pooling layers were not included in the architecture of the proposed convolutional neural network since they reduce the size of feature maps compared to the size of the original image and can decrease the segmentation performance of small psoriasis lesions and psoriasis lesions of complex shape.
 Results. The proposed convolutional neural network has been implemented in software and researched for solving the problem of psoriasis images segmentation.
 Conclusions. The use of the proposed convolutional neural network made it possible to enhance the segmentation performance of plaque and guttate psoriasis images, especially at the edges of the lesions. Prospects for further research are to study the performance of the proposed CNN then abrupt changes in color and illumination, blurring, as well as the complex background areas are present on dermatological images, for example, containing clothes or fragments of the interior. It is advisable to use the proposed CNN in other problems of color image processing to segment statistical or spectral-statistical texture regions on a uniform or textured background.

Machine Learning for Detection of Macroalgal Blooms in the Mar Menor Coastal Lagoon Using Sentinel-2

The Mar Menor coastal lagoon in southeastern Spain has experienced a decline in water quality due to increased nutrient input, leading to the eutrophication of the lagoon and the occurrence of microalgal and macroalgal blooms. This study analyzes the macroalgal bloom that occurred in the lagoon during the spring-summer of 2022. A set of machine learning techniques are applied to Sentinel-2 satellite imagery in order to obtain indicators of the presence of macroalgae in specific locations within the lagoon. This is supported by in situ observations of the blooming process in different areas of the Mar Menor. Our methodology successfully identifies the macroalgal bloom locations (accuracies above 98%, and Matthew’s Correlation Coefficients above 78% in all cases), and provides a probabilistic approach to understand the likelihood of occurrence of this event in given pixels. The analysis also identifies the key parameters contributing to the classification of pixels as algae, which could be used to develop future algorithms for detecting macroalgal blooms. This information can be used by environmental managers to implement early warning and mitigation strategies to prevent water quality deterioration in the lagoon. The usefulness of satellite observations for ecological and crisis management at local and regional scales is also highlighted.

HyFormer: Hybrid Transformer and CNN for Pixel-Level Multispectral Image Land Cover Classification.

To effectively solve the problems that most convolutional neural networks cannot be applied to the pixelwise input in remote sensing (RS) classification and cannot adequately represent the spectral sequence information, we propose a new multispectral RS image classification framework called HyFormer based on Transformer. First, a network framework combining a fully connected layer (FC) and convolutional neural network (CNN) is designed, and the 1D pixelwise spectral sequences obtained from the fully connected layers are reshaped into a 3D spectral feature matrix for the input of CNN, which enhances the dimensionality of the features through FC as well as increasing the feature expressiveness, and can solve the problem that 2D CNN cannot achieve pixel-level classification. Secondly, the features of the three levels of CNN are extracted and combined with the linearly transformed spectral information to enhance the information expression capability, and also used as the input of the transformer encoder to improve the features of CNN using the powerful global modelling capability of the Transformer, and finally the skip connection of the adjacent encoders to enhance the fusion between different levels of information. The pixel classification results are obtained by MLP Head. In this paper, we mainly focus on the feature distribution in the eastern part of Changxing County and the central part of Nanxun District, Zhejiang Province, and conduct experiments based on Sentinel-2 multispectral RS images. The experimental results show that the overall accuracy of HyFormer for the study area classification in Changxing County is 95.37% and that of Transformer (ViT) is 94.15%. The experimental results show that the overall accuracy of HyFormer for the study area classification in Nanxun District is 95.4% and that of Transformer (ViT) is 94.69%, and the performance of HyFormer on the Sentinel-2 dataset is better than that of the Transformer.

A new soil-plant image segmentation approach using K-means applied to experimental plots

The image segmentation procedure is fundamental in the phenotyping of plant images. Supervised algorithms have been used for pixel soil plant segmentation. Recent research has used the K-means algorithm to evaluate the segmentation of agronomic images in different crops with different databases. The algorithm has shown good performance in the pixel clustering process despite not being able to classify them directly. The present research intends to propose the use of the K-means algorithm in image segmentation and pixel classification in sugarcane images. 37,430-pixel samples referring to soil and vegetation were manually extracted from some images. This information was used to train and evaluate supervised models. The model with the best performance was considered the "standard" model. A rule that can serve as empirical support to interpret the clusters formed by K-means by assigning a label to each pixel was proposed. Then K-means was used to segment all images and classify the pixels. The vegetation index was used as features and the standard model classification was used as a true label. The measures recall, F1Score, precision, and accuracy were used as a performance measure of K-means, and the mask of each produced to compare the final result of the two approaches, highlighting the vegetation. Using K-means provided better-defined edges than Logistic Regression (standard model) and considerably distinguished the occurrence of soil between the leaves, with precision ranging from 0.77 to 0.92. These results expressed the importance of vegetation index to the clusterization process and showed that K-means ally to an interpretation clusters rule, which could be used to classify pixels in images.

Learning How to Detect Salient Objects in Nighttime Scenes

The detection of salient objects in nighttime scene settings is an essential research issue in computer vision. None of the known approaches can accurately anticipate salient objects in the nighttime scenes. Due to the lack of visible light, spatial visual information cannot be accurately perceived by traditional and deep network models. This paper proposed a Mountain Basin Network (MBNet) to identify salient objects for distinguishing the pixel-level saliency of low-light images. To improve the objects localizations and pixel classification performances, the proposed model incorporated a High-Low Feature Aggregation Module (HLFA) to synchronize the information from a high-level branch (named Bal-Net) and a low-level branch (called Mol-Net) to fuse the global and local context, and a Hierarchical Supervision Module (HSM) was embedded to aid in obtaining accurate salient objects, particularly the small ones. In addition, a multi-supervised integration technique was explored to optimize the structure and borders of salient objects. In the meantime, to facilitate more investigation into nighttime scenes and assessment of visual saliency models, we created a new nighttime dataset consisting of thirteen categories and a total of one thousand low-light images. Our experimental results demonstrated that the suggested MBNet model outperforms seven current state-of-the-art methods for salient object detection in nighttime scenes.

HMLNet: a hierarchical metric learning network with dual attention for change detection in high-resolution remote sensing images

ABSTRACT Change detection (CD) in high-resolution remote sensing images (RSIs) can be regarded as a binary visual recognition problem. Metric learning (ML) is a reliable method to determine pixel class attributes based on discriminative distance function between learnable image features. However, most related works learn discriminative distance functions in single-scale feature pairs, which suffer from slow convergence and poor local optima, partially due to that the loss function employs only large-scale feature samples while not interacting with the other scale features. Furthermore, more effective features are a prerequisite for improving the performance of ML-based CD. Hence, we propose a novel hierarchical metric learning network (HMLNet) with dual attention for CD in RSIs, where the key point is that hierarchical metric learning is performed in an ensemble manner to improve detection accuracy and accelerate model convergence. Specifically, based on the features extracted from the encoder-decoder backbone, we construct a feature pyramid to handle the complex details of objects at various scales in RSIs, and then perform metric learning between the paired pyramid features at the same scale. In addition, the dual-attention module is proposed to enhance the internal consistency of changed objects and effectively obtain more detailed information by acting on multi-scale pyramid features. Extensive experiments on the two public RSIs CD datasets, and the results demonstrate that the proposed HMLNet can accurately locate changed objects, which consistently outperforms the state-of-the-art CD competitors.

Land Cover Classification from Hyperspectral Images via Weighted Spatial–Spectral Joint Kernel Collaborative Representation Classifier

The continuous changes in Land Use and Land Cover (LULC) produce a significant impact on environmental factors. Highly accurate monitoring and updating of land cover information is essential for environmental protection, sustainable development, and land resource planning and management. Recently, Collaborative Representation (CR)-based methods have been widely used in land cover classification from Hyperspectral Images (HSIs). However, most CR methods consider the spatial information of HSI by taking the average or weighted average of spatial neighboring pixels of each pixel to improve the land cover classification performance, but do not take the spatial structure information for pixels into account. To address this problem, a novel Weighted Spatial–Spectral Joint CR Classification (WSSJCRC) method is proposed in this paper. WSSJCRC only performs spatial filtering on HSI through a weighted spatial filtering operator to alleviate the spectral shift caused by adjacency effect, but also utilizes the labeled training pixels to simultaneously represent each test pixel and its spatial neighborhood pixels to consider the spatial structure information of each test pixel to assist the classification of the test pixel. On this basis, the kernel version of WSSJCRC (i.e., WSSJKCRC) is also proposed, which projects the hyperspectral data into the kernel-induced high-dimensional feature space to enhance the separability of nonlinear samples. The experimental results on three real hyperspectral scenes show that the proposed WSSJKCRC method achieves the best land cover classification performance among all the compared methods. Specifically, the Overall Accuracy (OA), Average Accuracy (AA), and Kappa statistic (Kappa) of WSSJKCRC reach 96.21%, 96.20%, and 0.9555 for the Indian Pines scene, 97.02%, 96.64%, and 0.9605 for the Pavia University scene, and 95.55%, 97.97%, and 0.9504 for the Salinas scene, respectively. Moreover, the proposed WSSJKCRC method obtains the promising accuracy with OA over 95% on the three hyperspectral scenes under the situation of small-scale labeled samples, thus effectively reducing the labeling cost for HSI.

Road Network Detection from Aerial Imagery of Urban Areas Using Deep ResUNet in Combination with the B-snake Algorithm

Road network detection is critical to enhance disaster response and detecting a safe evacuation route. Due to expanding computational capacity, road extraction from aerial imagery has been investigated extensively in the literature, specifically in the last decade. Previous studies have mainly proposed methods based on pixel classification or image segmentation as road/non-road images, such as thresholding, edge-based segmentation, k-means clustering, histogram-based segmentation, etc. However, these methods have limitations of over-segmentation, sensitivity to noise, and distortion in images. This study considers the case study of Hawkesbury Nepean valley, NSW, Australia, which is prone to flood and has been selected for road network extraction. For road area extraction, the application of semantic segmentation along with residual learning and U-Net is suggested. Public road datasets were used for training and testing purposes. The study suggested a framework to train and test datasets with the application of the deep ResUnet architecture. Based on maximal similarity, the regions were merged, and the road network was extracted with the B-snake algorithm application. The proposed framework (baseline + region merging + B-snake) improved performance when evaluated on the synthetically modified dataset. It was evident that in comparison with the baseline, region merging and addition of the B-snake algorithm improved significantly, achieving a value of 0.92 for precision and 0.897 for recall.

Metal artefact reduction in digital tomosynthesis using component decomposition

We propose a technique for metal artefact reduction in digital tomosynthesis reconstruction. Although the problem was addressed earlier in the literature, we suggest another approach, which is, in our opinion, simpler, and easier to implement. It is a two-stage algorithm. At the first stage, attenuation images are segmented by decomposing their intensity distributions into gaussian-like components. Statistical information contained in each component is used for pixel classification. Components corresponding to metallic objects are identified, and a pixel threshold value separating regions occupied by metal objects from the rest of the image is found. Based on this value, at the second stage, a smooth mapping of image intensity is applied. This makes dense regions transparent, resulting in the artefact reduction in reconstruction. The methodology is demonstrated by several examples.

Transfer learning for real-time crater detection on asteroids using a Fully Convolutional Neural Network

This paper proposes the use of Transfer Learning from the Moon to Ceres for real-time autonomous crater detection operations on asteroids. This approach, based on the use of a U-Net for crater segmentation, starts with the network training on a Moon Digital Elevation Model; the trained model is then refined using two different strategies, one based on data augmentation and one based on fine tuning with small sets of Ceres images. Data from the DAWN mission was used, while the reference crater catalogue for Ceres is the one created by Zeilnhofer from LAMO data, containing 44,594 craters with diameter greater than 1 km in a latitude range of 84.66°S–89.62°N, full longitude. The loss function chosen during network training is the Focal Tversky Loss, which takes into consideration the imbalanced distribution of class pixels in the ground truth (pixels corresponding to crater rims represent approximately only 3% over the whole dataset). For both the Moon and Ceres, the U-Net output is post-processed with a template matching algorithm. The detected craters are counted and compared to the crater catalogue locations in order to compute metrics performances (precision, recall, F1 score) in the crater domain. Post-processing on the Moon reached performance of 84.12% F1 score on the baseline model. The baseline and augmented models gave a lower performance on Ceres data, although comparable with the outcomes found in other works, with a best F1 score of 70.17%. However, when the model is completely retrained starting from the baseline, it ends up with a best F1 score of 79.96%. Finally, a strategy for the actual on-board implementation is proposed considering the NVIDIA Jetson TX2 accelerator. We demonstrated that the inference time of the model on the accelerator is on the order of few milliseconds without loss of performances, thus proving the possibility to consider this approach for future real-time on-board navigation systems.

A Combined Region- and Pixel-Based Deep Learning Approach for Quantifying Abdominal Adipose Tissue in Adolescents Using Dixon Magnetic Resonance Imaging.

The development of adipose tissue during adolescence may provide valuable insights into obesity-associated diseases. We propose an automated convolutional neural network (CNN) approach using Dixon-based magnetic resonance imaging (MRI) to quantity abdominal subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT) in children and adolescents. 474 abdominal Dixon MRI scans of 136 young healthy volunteers (aged 8-18) were included in this study. For each scan, an axial fat-only Dixon image located at the L2-L3 disc space and another image at the L4-L5 disc space were selected for quantification. For each image, an outer and an inner region around the abdomen wall, as well as SAT and VAT pixel masks, were generated by expert readers as reference standards. A standard U-Net CNN architecture was then used to train two models: one for region segmentation and one for fat pixel classification. The performance was evaluated using the dice similarity coefficient (DSC) with fivefold cross-validation, and by Pearson correlation and the Student's t-test against the reference standards. For the DSC results, means and standard deviations of the outer region, inner region, SAT, and VAT comparisons were 0.974 ± 0.026, 0.997 ± 0.003, 0.981 ± 0.025, and 0.932 ± 0.047, respectively. Pearson coefficients were 1.000 for both outer and inner regions, and 1.000 and 0.982 for SAT and VAT comparisons, respectively (all p = NS). These results show that our method not only provides excellent agreement with the reference SAT and VAT measurements, but also accurate abdominal wall region segmentation. The proposed combined region- and pixel-based CNN approach provides automated abdominal wall segmentation as well as SAT and VAT quantification with Dixon MRI and enables objective longitudinal assessment of adipose tissues in children during adolescence.

Automatic retrieval of volcanic SO2 emission source from TROPOMI products

Volcanic sulfur dioxide (SO2) satellite observations are key for monitoring volcanic activity, and for mitigation of the associated risks on both human health and aviation safety. Automatic analysis of this data source, including robust source emission retrieval, is in turn essential for near real-time monitoring applications. We have developed fast and accurate SO2 plume classifier and segmentation algorithms using classic clustering, segmentation and image processing techniques. These algorithms, applied to measurements from the TROPOMI instrument onboard the Sentinel-5 Precursor platform, can help in the accurate source estimation of volcanic SO2 plumes originating from various volcanoes. In this paper, we demonstrate the ability of different pixel classification methodologies to retrieve SO2 source emission with a good accuracy. We compare the algorithms, their strengths and shortcomings, and present plume classification results for various active volcanoes throughout the year 2021, including examples from Etna (Italy), Sangay and Reventador (Ecuador), Sabancaya and Ubinas (Peru), Scheveluch and Klyuchevskoy (Russia), as well as Ibu and Dukono (Indonesia). The developed algorithms, shared as open-source code, contribute to improving analysis and monitoring of volcanic emissions from space.

Improvement of pixel classification by the simultaneous use of spectral and spatial information in the framework of spectroscopic imaging

Hyperspectral imaging technology is developing in a very fast way. We find it today in many analytical developments using different spectroscopies for sample classification purposes. Instrumental developments allow us to acquire more and more data in shorter and shorter periods of time while improving their quality. Therefore, we are going in the right direction as far as the measure is concerned. On the other hand, we can make a more mixed assessment for the hyperspectral imaging data processing. Indeed, the data acquired in spectroscopic imaging have the particularity of encoding both spectral and spatial information. Unfortunately, in chemometrics, almost all classification approaches today only use spectral information from three-dimensional hyperspectral data arrays. To be more precise, an approach encompassing the unfolding/refolding of such arrays is often applied beforehand because the majority of algorithms for analysing these data are not capable of handling them in their original structure. Spatial information is therefore lost during the chemometric exploration. The study of the spectral part of the acquired data array alone is clearly a limitation that we propose to overcome in this work. 2-D Stationary Wavelet Transform will be used in the data preprocessing phase to ensure the joint use of spectral and spatial information. Two spectroscopic datasets will then be used to evaluate the potential of our approach in the context of supervised classification.

MATNet: A Combining Multi-Attention and Transformer Network for Hyperspectral Image Classification

Hyperspectral image (HSI) has rich spatial-spectral information, high spectral correlation and large redundancy between information. Due to the sparse background distribution of HSI, existing methods generally perform poorly for the classification of class pixels located in the boundary areas of land cover categories. This is largely because the network is vulnerable to surrounding redundant information during the training stage, leading to inaccurate feature extraction and thus poor generalization ability of the model. Based on previous work, we propose a HSI classification network called MATNet which combines multi-attention and Transformer. The network first uses spatial attention and channel attention to pay more attention to the more significant information parts, then uses tokenizer module to make a semantic level representation of different categories of ground objects, and then performs deep semantic feature extraction using the transformer encoder module. Finally, we design a loss function called Lpoly, which adds a polynomial to the label smoothing loss to tune the original first polynomial to accommodate different datasets and tasks. We perform experiments in several well-known HSI datasets as well as for visualization. The results show that our proposed MATNet performs well in extracting spatial-spectral features of HSIs as well as understanding semantic degrees of semantic degrees.

Liver Tumors Segmentation Using 3D SegNet Deep Learning Approach

An ultrasonic filter detects signs of malignant tumors by analysing the image’s pixel quality fluctuations caused by a liver ailment. Signs of malignant growth proximity are identified in an ultrasound filter through image pixel quality variations from a liver’s condition. Those changes are more common in alcoholic liver conditions than in other etiologies of cirrhosis, suggesting that the cause may be alcohol instead of liver disease. Existing Two-Dimensional (2D) ultrasound data sets contain an accuracy rate of 85.9% and a 2D Computed Tomography (CT) data set of 91.02%. The most recent work on designing a Three-Dimensional (3D) ultrasound imaging system in or close to real-time is examined. In this article, a Deep Learning (DL) model is implemented and modified to fit liver CT segmentation, and a semantic pixel classification of road scenes is recommended. The architecture is called semantic pixel-wise segmentation and comprises a hierarchical link of encoder-decoder layers. A standard data set was used to test the proposed model for liver CT scans and the tumor accuracy in the training phase. For the normal class, we obtained 100% precision for chronic cirrhosis hepatitis (73%), offset cirrhosis (59.26%), and offensive cirrhosis (91.67%) for chronic hepatitis or cirrhosis (73,0%). The aim is to develop a Computer-Aided Detection (CAD) screening tool to detect steatosis. The results proved 98.33% exactness, 94.59% sensitivity, and 92.11% case with Convolutional Neural Networks (CNN) classification. Although the classifier’s performance did not differentiate so clearly at this level, it was recommended that CNN generally perform better due to the good relationship between Area under the Receiver Operating Characteristics Curve (AUC) and accuracy.

Crack Segmentation Based on Fusing Multi-Scale Wavelet and Spatial-Channel Attention

Accurate and reliable crack segmentation is a challenge and meaningful task. In this article, aiming at the characteristics of cracks on the concrete images, the intensity frequency information of source images which is obtained by Discrete Wavelet Transform (DWT) is fed into deep learning-based networks to enhance the ability of network on crack segmentation. To well integrate frequency information into network an effective and novel DWTA module based on the DWT and scSE attention mechanism is proposed. The semantic information of cracks is enhanced and the irrelevant information is suppressed by DWTA module. And the gap between frequency information and convolution information from network is balanced by DWTA module which can well fuse wavelet information into image segmentation network. The Unet-DWTA is proposed to preserved the information of crack boundary and thin crack in intermediate feature maps by adding DWTA module in the encoder-decoder structures. In decoder, diverse level feature maps are fused to capture the information of crack boundary and the abstract semantic information which is beneficial to crack pixel classification. The proposed method is verified on three classic datasets including CrackDataset, CrackForest, and DeepCrack datasets. Compared with the other crack methods, the proposed Unet-DWTA shows better performance based on the evaluation of the subjective analysis and objective metrics about image semantic segmentation.

Hyper-Spectral Image Pixel Classification Based on Golden Sine and Chaotic Spotted Hyena Optimization Algorithm

Hyper-Spectral Image Pixel Classification Based on Golden Sine and Chaotic Spotted Hyena Optimization Algorithm