Network SegNet Research Articles

IScan: Detection of Colorectal Cancer from CT Scan Images Using Deep Learning

Colorectal cancer, a highly lethal form of cancer, can be treated effectively if detected early. However, the current diagnosis process involves a time-consuming and manual review of CT scans to identify cancerous regions and behavior, leading to resource consumption, subjectivity, and dependency on manual assessment. We propose a 3-phase deep neural system for automated colorectal cancer detection using CT scan images to address these challenges. It includes a SegNet network to identify tumor locations, an InceptionResNet V2 network to classify tumors as benign or malignant, and an analysis of tumor area cum perimeter to predict the cancer stage. The proposed model offers a fully automated solution by combining these functionalities under a single umbrella. In real-life CT scans from 37 patients, the proposed model achieved 95.8% ROI segmentation accuracy, a dice coefficient of 0.6214, 69.75% IoU score, and 95.83% tumor classification accuracy. The unique approach using Radial Length (RL) and Circularity (C) parameters predicted the T-stage with close to 85% accuracy. Based on these outcomes, the proposed system establishes itself as a reliable and suitable alternative to traditional cancer diagnosis techniques by leveraging the power of automation, deep learning, and innovative parameter analysis.

Classifying alzheimer's disease from SMRI data using a hybrid deep learning approaches

The chance of developing "Alzheimer's Disease (AD)" increases every 5 years after 65 years of age, making it a particularly common form of neurodegenerative disorder among the older population. The use of "Magnetic Resonance Imaging (MRI)" to diagnose AD has grown in popularity in recent years. A further benefit of MRI is that it provides excellent contrast and exquisite structural detail. As a result, some studies have used biological markers backed by "structural MRI (sMRI)" to separate AD populations, which indicate differences in brain tissue size and degradation of the nervous system. The lack of properly segmented regions and essential features by the existing models might affect classification accuracy for AD. The categorization of AD in this study is based on sMRI. In this research, the hybrid Deep-Learning Models "SegNet and ResNet (SegResNet)" have been proposed for segmentation, feature extraction, and to classify the AD. SegNet network is used to identify and segment specific brain areas. Edges and circles are the SegNet's first levels, whereas the deeper layers acquire more nuanced and useful features. SegNet's last deconvolution layer produces a wide range of segmented images linked to the 3 categorization labels "Cognitive Normal (CN)", "Mild Cognitive Impairment (MCI)", and "AD" which the machine has earlier found out. To increase classification performance, the attributes of each segmented sMRI image serve as strong features of the labels. To enhance the feature information used for classification, a feature vector is built by combining the values of the pixel intensity of the segmented sMRI images. ResNet-101 classifiers are then used for characterizing vectors to identify the presence or absence of AD or MCI in each sMRI image. In terms of detection and classification accuracy, the proposed SegResNet Model is superior to the existing KNN, EFKNN, AANFIS, and ACS approaches

Research on road surface crack detection based on SegNet network

To enhance the precision and reliability of road crack detection, this study introduces an innovative neural network architecture. Strategies were implemented to effectively address the issue of overfitting resulting from the intricacy of the proposed SegCrackNet. Dropout layers, multi-level output fusion, and T-bridge block structures are employed in the network. This optimization allows for a more comprehensive exploitation of contextual information, demonstrating its instrumental role in the efficient detection of subtle variations. Experimental findings clearly demonstrate substantial improvements when compared to other network models. On the Crack500, Crack200, and pavement images datasets, remarkable enhancements in the average Intersection over Union (IoU) scores were observed, with increases of 4.3%, 9.4%, and 3.7%, respectively.

SegNet Network Architecture for Deep Learning Image Segmentation and Its Integrated Applications and Prospects

Semantic image segmentation is a crucial task in computer vision, with applications ranging from autonomous driving to medical image analysis. In recent years, deep learning has revolutionized this field, leading to the development of various neural network models aimed at improving segmentation accuracy. One such architecture is SegNet, which we explore in this article.SegNet's architecture consists of an encoder network, a corresponding decoder network, and a pixel-wise classification layer. The encoder network, resembling VGG16 with 13 convolutional layers, extracts high-level features from input images. The innovation lies in the decoder network's approach to upsampling, utilizing pooled indices from the encoder's maximum pooling step to perform non-linear up sampling. This eliminates the need for additional learning during up sampling, making SegNet efficient in both storage and computation.SegNet represents an exciting advancement in deep learning image segmentation. Its efficient architecture, memory-conscious design, and potential for real-time applications make it a valuable tool in the field of computer vision with promising integrated applications and prospects.

Bimodal SegNet: Fused instance segmentation using events and RGB frames

Object segmentation enhances robotic grasping by aiding object identification. Complex environments and dynamic conditions pose challenges such as occlusion, low light conditions, motion blur and object size variance. To address these challenges, we propose a Bimodal SegNet that fuses two types of visual signals, event-based data and RGB frame data. The proposed Bimodal SegNet network has two distinct encoders — one for RGB signal input and another for Event signal input, in addition to an Atrous Pyramidal Feature Amplification module. Encoders capture and fuse the rich contextual information from different resolutions via a Cross-Domain Contextual Attention layer while the decoder obtains sharp object boundaries. The evaluation of the proposed method undertakes five unique image degradation challenges including occlusion, blur, brightness, trajectory and scale variance on the Event-based Segmentation (ESD) Dataset. The results show a 4%–6% MIOU score improvement over state-of-the-art methods in terms of mean intersection over the union and pixel accuracy. The source code, dataset and model are publicly available at: https://github.com/sanket0707/Bimodal-SegNet.

DANet: A Semantic Segmentation Network for Remote Sensing of Roads Based on Dual-ASPP Structure

Semantic segmentation of roads in remote-sensing images is a challenging task. This paper proposes a semantic segmentation model, DANet, for remote-sensing image road semantic segmentation. The model addresses the problems of missing, misclassification, and strong jaggedness of segmented target edges faced by other semantic segmentation networks when dealing with complex and diverse remote-sensing images. The proposed model uses two ASPP structures for multi-scale feature fusion and combines the DarkNet network structure for downsampling with the SegNet network structure for upsampling. This improves the model’s ability to extract road feature information from remote-sensing images. Using the CHN–CUG Roads Dataset, we have confirmed that the proposed network structure, Re, has demonstrated a 1.15% improvement in accuracy compared to U-Net. Furthermore, the road IoU has shown a 1.09% enhancement compared to HRNet-V2. Additionally, there is a 1.13% increase in F1-score compared to U-Net.

Increasing the localization accuracy of visual SLAM with semantic segmentation and motion consistency detection in dynamic scenes1

Visual Simultaneous Localization and Mapping (SLAM) is one of the key technologies for intelligent mobile robots. However, most of the existing SLAM algorithms have low localization accuracy in dynamic scenes. Therefore, a visual SLAM algorithm combining semantic segmentation and motion consistency detection is proposed. Firstly, the RGB images are segmented by SegNet network, the prior semantic information is established and the feature points of high-dynamic objects are removed; Secondly, motion consistency detection is carried out, the fundamental matrix is calculated by the improved Random Sample Consistency (RANSAC) algorithm, the abnormal feature points are output by the epipolar geometry method, and the feature points of low-dynamic objects are eliminated by combining the prior semantic information. Thirdly, the static feature points are used for pose estimation. Finally, the proposed algorithm is tested on the TUM dataset, the algorithm in this paper reduces the average RMSE of ORB-SLAM2 by 93.99% in highly dynamic scenes, which show that the algorithm can effectively improve the localization accuracy of the visual SLAM system in dynamic scenes.

IMFSegNet: Cost-effective and objective quantification of intramuscular fat in histological sections by deep learning

The assessment of muscle condition is of great importance in various research areas. In particular, evaluating the degree of intramuscular fat (IMF) in tissue sections is a challenging task, which today is still mostly performed qualitatively or quantitatively by a highly subjective and error-prone manual analysis. We here realize the mission to make automated IMF analysis possible that (i) minimizes subjectivity, (ii) provides accurate and quantitative results quickly, and (iii) is cost-effective using standard hematoxylin and eosin (H&E) stained tissue sections. To address all these needs in a deep learning approach, we utilized the convolutional encoder-decoder network SegNet to train the specialized network IMFSegNet allowing to accurately quantify the spatial distribution of IMF in histological sections. Our fully automated analysis was validated on 17 H&E-stained muscle sections from individual sheep and compared to various state-of-the-art approaches. Not only does IMFSegNet outperform all other approaches, but this neural network also provides fully automated and highly accurate results utilizing the most cost-effective procedures of sample preparation and imaging. Furthermore, we shed light on the opacity of black-box approaches such as neural networks by applying an explainable artificial intelligence technique to clarify that the success of IMFSegNet actually lies in identifying the hard-to-detect IMF structures. Embedded in our open-source visual programming language JIPipe that does not require programming skills, it can be expected that IMFSegNet advances muscle condition assessment in basic research across multiple areas as well as in research fields focusing on translational clinical applications.

A hierarchical visual model for robot automatic arc welding guidance

PurposeThe traditional vision system cannot automatically adjust the feature point extraction method according to the type of welding seam. In addition, the robot cannot self-correct the laying position error or machining error. To solve this problem, this paper aims to propose a hierarchical visual model to achieve automatic arc welding guidance.Design/methodology/approachThe hierarchical visual model proposed in this paper is divided into two layers: welding seam classification layer and feature point extraction layer. In the welding seam classification layer, the SegNet network model is trained to identify the welding seam type, and the prediction mask is obtained to segment the corresponding point clouds. In the feature point extraction layer, the scanning path is determined by the point cloud obtained from the upper layer to correct laying position error. The feature points extraction method is automatically determined to correct machining error based on the type of welding seam. Furthermore, the corresponding specific method to extract the feature points for each type of welding seam is proposed. The proposed visual model is experimentally validated, and the feature points extraction results as well as seam tracking error are finally analyzed.FindingsThe experimental results show that the algorithm can well accomplish welding seam classification, feature points extraction and seam tracking with high precision. The prediction mask accuracy is above 90% for three types of welding seam. The proposed feature points extraction method for each type of welding seam can achieve sub-pixel feature extraction. For the three types of welding seam, the maximum seam tracking error is 0.33–0.41 mm, and the average seam tracking error is 0.11–0.22 mm.Originality/valueThe main innovation of this paper is that a hierarchical visual model for robotic arc welding is proposed, which is suitable for various types of welding seam. The proposed visual model well achieves welding seam classification, feature point extraction and error correction, which improves the automation level of robot welding.

Generating and Restoring Private Face Images for Internet of Vehicles Based on Semantic Features and Adversarial Examples

The disclosure of face image features can seriously threaten the security of user information, which limits the application of face recognition technology in Internet of Vehicles. This paper proposes a new method of generating and restoring private face images based on semantic features and adversarial examples. The Segnet network first segments the face images semantically, then a generative adversarial network generates adversarial examples and perturbs the semantic features of the face image. The perturbation position can be accurately controlled through a coefficient matrix as the identity tag of the face image is concealed steganographically. A restoration network is trained to extract the real identity tag from the private face image using a discriminator against the generation network, then it restores the private face image to its original state. Compared to other state-of-the-art methods, private face images generated by the proposed method experimentally show high detection resistance, better quality, and stronger median filtering defense.

Application and Research of the Image Segmentation Algorithm in Remote Sensing Image Buildings

Aiming at the problems of low building segmentation accuracy and blurred edges in high-resolution remote sensing images, an improved fully convolutional neural network is proposed based on the SegNet network. First, GELU, which performs well in deep learning tasks, is selected as the activation function to avoid neuron deactivation. Second, the improved residual bottleneck structure is used in the encoding network to extract more building features. Then, skip connections are used to fuse images The low-level and high-level semantic features are used to assist image reconstruction. Finally, an improved edge correction module is connected at the end of the decoding network to further correct the edge details of the building and improve the edge integrity of the building. Experiments are carried out on the Massachusetts building dataset, and the precision rate, recall rate, and F1 value reach 93.5%, 79.3%, and 81.9%, respectively, and the comprehensive evaluation index F1 value is improved by about 5% compared with the basic network.

Image Segmentation of a Sewer Based on Deep Learning

An accurate assessment of the type and extent of sewer damage is an important prerequisite for maintenance and repair. At present, distinguishing drainage pipe defect types in the engineering field mainly relies on the human eye, which is time consuming, labor intensive, and subjective. Some studies have used deep learning to classify the types of pipe defects, but this method can only identify one main pipe defect. However, sometimes a combination of defects, such as corrosion and precipitation on a section of pipe wall, can be classified as one category by picture classification, which is significantly different from the reality. Furthermore, the deep learning method for defect classification is unable to pinpoint the precise location and severity of a defect or estimate the number of flaws and the cost of maintenance and repair. Therefore, an image segmentation method based on deep convolutional neural networks is proposed to achieve pixel-level image segmentation of defect regions while classifying pipe defects. Compared with the deep learning network for defect classification, it can segment a variety of defects and reduce the number of samples, which is convenient for defect measurement. First, the image defect locations of seven typical defects were manually labeled to create the dataset. Then, a model based on the SegNet network was used to label defect areas automatically in an image. The pipeline image dataset was used to test the previously trained model using the CamVid dataset. Finally, the model was applied to drainage pipe network images that were provided by periscope and closed-circuit television inspection cameras, and the pixel accuracy of image segmentation reached 80%. From the results, it can be concluded that image segmentation and annotation technology based on deep learning is applicable to sewer defect detection. The identification results of pipeline defects were accurate. The SegNet model is a reliable method for image analysis of pipeline defects, which can accurately evaluate the type and degree of sewer damage.

Improved Feature Pyramid Convolutional Neural Network for Effective Recognition of Music Scores

Music written by composers and performed by multidimensional instruments is an art form that reflects real-life emotions. Historically, people disseminated music primarily through sheet music recording and oral transmission. Among them, recording music in sheet music form was a great musical invention. It became the carrier of music communication and inheritance, as well as a record of humanity's magnificent music culture. The advent of digital technology solves the problem of difficult musical score storage and distribution. However, there are many drawbacks to using data in image format, and extracting music score information in editable form from image data is currently a challenge. An improved convolutional neural network for musical score recognition is proposed in this paper. Because the traditional convolutional neural network SEGNET misclassifies some pixels, this paper employs the feature pyramid structure. Use additional branch paths to fuse shallow image details, shallow texture features that are beneficial to small objects, and high-level features of global information, enrich the multi-scale semantic information of the model, and alleviate the problem of the lack of multiscale semantic information in the model. Poor recognition performance is caused by semantic information. By comparing the recognition effects of other models, the experimental results show that the proposed musical score recognition model has a higher recognition accuracy and a stronger generalization performance. The improved generalization performance allows the musical score recognition method to be applied to more types of musical score recognition scenarios, and such a recognition model has more practical value.

Image Segmentation under the Optimization Algorithm of Krill Swarm and Machine Learning.

This study aims to improve the efficiency and accuracy of image segmentation, and to compare and study traditional threshold-based image segmentation methods and machine learning model-based image segmentation methods. The krill herb optimization algorithm is combined with the traditional maximum between-class variance function to form a new graph segmentation algorithm. The pet dataset is used to train the algorithm model and build an image semantic segmentation system. The results show that when the traditional Ostu algorithm performs image single-threshold segmentation, the number of iterations is about 256. When double-threshold segmentation is performed, the number of iterations increases exponentially, and the execution time is about 2 s. The number of iterations of the improved Krill Herd algorithm in single-threshold segmentation is 6.95 times, respectively. The execution time for double-threshold segmentation is about 0.24 s. The number of iterations is only improved by a factor of 0.19. The average classification accuracy of the Unet network model and the SegNet network model is 86.3% and 91.9%, respectively. The average classification accuracy of the DC-Unet network model reaches 93.1%. This shows that the proposed fusion algorithm has high optimization efficiency and stronger practicability in multithreshold image segmentation. The DC-Unet network model can improve the image detail segmentation effect. The research provides a new idea for finding an efficient and accurate image segmentation method.

An improved SegNet network model for accurate detection and segmentation of car body welding slags

An improved SegNet network model for accurate detection and segmentation of car body welding slags

Precise Segmentation of Remote Sensing Cage Images Based on SegNet and Voting Mechanism

HighlightsA Remote Sensing Cage Segmentation (RSCS) dataset is constructed.The SegNet network is introduced to achieve precise segmentation.Random cropping, data augmentation, and three-channel separation operations are used to construct the training sets.The proposed sliding window overlap cropping method and two rounds of voting are used to improve the segmentation accuracy.Abstract. In mariculture, improper cage layout and excessive density of mariculture will lead to deterioration of water quality and the growth of harmful bacteria. However, relying solely on manual measurement will consume a considerable amount of manpower and material resources. Therefore, we propose a precise segmentation scheme for remote sensing cage images based on SegNet and voting mechanism. First, a Remote Sensing Cage Segmentation (RSCS) dataset is constructed. Second, the number of collected samples is too small and the sample sizes are too large. Random cropping, data augmentation, and three-channel separation operations are used to construct the training sets. Nine training sets consisting of three image sizes and three single channels are generated. Finally, the proposed sliding window overlap cropping method and two rounds of voting are used on the test samples to improve the segmentation accuracy. The experimental results show that using sliding window overlap cropping, three-channel voting, and three-size voting can improve mIoU (mean Intersection over Union) by up to 0.9%, 1.9%, and 0.6%, respectively. By using the proposed final scheme, the mIoU of test samples can reach 0.89. Keywords: Mariculture, Remote image segmentation, SegNet, Sliding window overlap cropping, Voting mechanism.

Segnet Network Algorithm-Based Ultrasound Images in the Diagnosis of Gallbladder Stones Complicated with Gallbladder Carcinoma and the Relationship between P16 Expression with Gallbladder Carcinoma.

The study focused on how to improve the diagnostic coincidence rate of patients with gallbladder stones and gallbladder cancer based on an optimized Segnet network algorithm and the relationship of gallbladder cancer with multiple tumor suppressor 1 (P16). 300 patients diagnosed with gallbladder cancer in the hospital were selected as the research subjects. The pyramid pooling operation was incorporated into the original Segnet network algorithm, and its performance was evaluated, factoring into the intersection of union (IoU), algorithm precision (Pre), and recall rate (Recall). After 8 hours of fasting, conventional ultrasound and contrast-enhanced ultrasound examinations were performed, and the images were evaluated by three experienced ultrasound diagnosticians. The positive signal of P16 immunohistochemical staining was brownish yellow, which was generally concentrated in the nucleus, and a small part was located in the cytoplasm. In each slice, ten visual fields were selected. Then, they were observed under a high-power mirror, and the number was counted. It was found that the optimized Segnet network algorithm increased the IoU by 7.3%, the precision by 8.2%, and the recall rate by 11.1%. The diagnostic coincidence rates of conventional ultrasound and contrast-enhanced ultrasound examinations for gallbladder cancer were 78.13% (25/32) and 87.5% (25/32), respectively. The positive expression rate of P16 in gallbladder adenocarcinoma (47.06%) was significantly lower than that of acute cholecystitis with gallbladder stones (84.38%) and gallbladder polyps (67.16%) (P < 0.05). The positive expression rate of P16 in patients with stage III and stage IV (33.33% and 40%) was significantly lower than that in patients with stages I and II (87.5% and 80%) (P < 0.05). The positive expression rate of P16 in high differentiation (86.67%) was significantly higher than that of moderate differentiation (40%) and poor differentiation (28.57%) (P < 0.05). In short, contrast-enhanced ultrasound can effectively improve the diagnostic coincidence rate of gallbladder cancer, and the expression of P16 in gallbladder cancer is closely related to tumor staging and differentiation.

Use of convolutional neural networks with encoder-decoder structure for predicting the inverse operator in hydraulic tomography

In this manuscript, we discuss the capabilities of a deep learning algorithm implemented with the Conventional Neural Network concept to characterize the hydraulic properties of aquifers. The algorithm called CNN-HT is designed to predict the inverse operator of hydraulic tomography using a synthetic training dataset in which the hydraulic head data associated with pumping tests are linked to hydraulic transmissivity field. This approach relies on an adaptation of the SegNet network that was initially developed to process image segmentation. The SegNet is composed of encoders and decoders networks. In the encoder, sequential operations with multiple filters, as convolution, batch normalization, max-pooling are performed to identify feature maps of the input data. In the decoder, the up-sampling, convolution, batch normalization and regression operations are used to prepare the output by recovering the loss of spatial resolution that occurred in the encoder process. In this adaptation, we used the least-square iterative formulation at the initial iteration with Jacobian matrix to resize the hydraulic head data to match the size of the output (transmissivity field). This protocol was applied to the hydraulic head data computed numerically by solving the groundwater flow equation for a given transmissivity field, generated geostatistically with Gaussian and spherical variograms. A part of this data was used for training the network and the other part to test its performance. The test step confirmed the effectiveness of this tool in reconstructing the main heterogeneities of the hydraulic properties, and its effectiveness is related to the nature and quantity of the training data. Moreover, the CNN-HT method provided inversion results of the same quality than those obtained with the Gauss-Newton algorithm using the finite difference or adjoint state method in the computation of the Jacobian matrix. However, the computational time is longer in CNN-HT but this time can be less or of the same order as that of Gauss-Newton using finite difference method.

Deep Learning Semantic Segmentation for Water Level Estimation Using Surveillance Camera

The interest in visual-based surveillance systems, especially in natural disaster applications, such as flood detection and monitoring, has increased due to the blooming of surveillance technology. In this work, semantic segmentation based on convolutional neural networks (CNN) was proposed to identify water regions from the surveillance images. This work presented two well-established deep learning algorithms, DeepLabv3+ and SegNet networks, and evaluated their performances using several evaluation metrics. Overall, both networks attained high accuracy when compared to the measurement data but the DeepLabv3+ network performed better than the SegNet network, achieving over 90% for overall accuracy and IoU metrics, and around 80% for boundary F1 score (BF score), respectively. When predicting new images using both trained networks, the results show that both networks successfully distinguished water regions from the background but the outputs from DeepLabv3+ were more accurate than the results from the SegNet network. Therefore, the DeepLabv3+ network was used for practical application using a set of images captured at five consecutive days in the study area. The segmentation result and water level markers extracted from light detection and ranging (LiDAR) data were overlaid to estimate river water levels and observe the water fluctuation. River water levels were predicted based on the elevation from the predefined markers. The proposed water level framework was evaluated according to Spearman’s rank-order correlation coefficient. The correlation coefficient was 0.91, which indicates a strong relationship between the estimated water level and observed water level. Based on these findings, it can be concluded that the proposed approach has high potential as an alternative monitoring system that offers water region information and water level estimation for flood management and related activities.

Research on Image Super-resolution Technology Based on Sparse Constraint SegNet Network

Aiming at the problems of long time-consuming and low accuracy in extracting buildings with traditional machine learning methods. In this paper, the SegNet semantic segmentation models on deep learning is used to improve the algorithm, and a high-resolution remote sensing image building extraction algorithm based on sparse constrained SegNet are proposed. First, regular terms and Dropout are added to the SegNet model, which greatly reduces the occurrence of model over-fitting; secondly, in order to extract richer semantic features for the model, the algorithm introduces a pyramid pooling module; finally, the Lorentz function sparse constraint factor is introduced to the SPNet model, Construct a new semantic segmentation model LSPNet. In order to verify the reliability and applicability of the proposed algorithm, the optimized LSPNet model is used to identify and extract the buildings in the high-resolution data set. Experimental results show that compared with traditional machine learning methods, this method has the advantages of fast convergence and high accuracy, and has a good application prospect.