- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
PolSAR Image Semantic Segmentation Based on Deep Transfer Learning--Realizing Smooth Classification With Small Training Sets
摘要： Suffering from speckle noise and complex scattering phenomena, classification results of SAR images are usually noisy and shattered, which makes them difficult to use in practical applications. Deep-learning-based semantic segmentation realizes segmentation and categorization at the same time, and thus can obtain smooth and fine-grained classification maps. However, this kind of methods require large data sets with pixel-wise categorical annotations, which are time consuming and tedious to retrieve. Compared with photographs and optical remote sensing images, manually annotating SAR data is even harder, which results in a delay of using relevant techniques in this field. In this letter, a new data set is proposed to support semantic segmentation for high-resolution PolSAR images. Limited by the aforementioned problems, the data set is only a small one with 50 image patches. Therefore, two transfer learning strategies are proposed, which adopt the fully convolutional network (FCN) and U-net architecture, respectively, and use distinct pretraining data sets to adapt to different situations. The experiments demonstrate the good performance of both methods and a promising applicability of using small training sets. Moreover, although trained with small patches, both networks can perfectly apply on large images. The new data set and methods are hopeful to support various PolSAR applications as baselines.
关键词： polarimetry,SAR,image classification,Deep learning,image segmentation
更新于2025-09-23 15:22:29
-
[IEEE 2018 OCEANS - MTS/IEEE Kobe Techno-Ocean (OTO) - Kobe, Japan (2018.5.28-2018.5.31)] 2018 OCEANS - MTS/IEEE Kobe Techno-Oceans (OTO) - Combining Adaptive Thresholding and Region Filling for Xylene Spills Detection from Ultraviolet Images
摘要： Numerous marine chemical spill accidents have caused enormous damage to the marine ecological environment. Multiple researchers are engaged in providing effective method to detect chemical spill detection. In this paper, we compare the properties between ultraviolet (UV) images and visible images. It turns out that the characteristics of UV images help eliminate the background influence. Then, we develop a new algorithm for segmenting chemical spills from UV imagery. This algorithm, combining adaptive thresholding and region filling, can effectively solve the problem of uneven illumination than the Otsu and FCM algorithms. Moreover, this algorithm does well in time consuming. Experimental results on UV imagery demonstrate that our approach can accurately segment chemical spill without producing too much false alarms.
关键词： region filling,UV images,segmentation,adaptive thresholding,chemical spill
更新于2025-09-23 15:22:29
-
[IEEE 2018 3rd International Conference on Pattern Analysis and Intelligent Systems (PAIS) - Tebessa, Algeria (2018.10.24-2018.10.25)] 2018 3rd International Conference on Pattern Analysis and Intelligent Systems (PAIS) - A system for the automatic detection of glaucoma using retinal images
摘要： Glaucoma is an optic neuropathy and it principal cause of blindness in the world. In this paper, a system able to treat and analyze the Visual Field (VF) images and Optical Coherence Tomography of the Ganglion Cell Layer (OCT-GCL) images is proposed, in order to help early detection of glaucoma in its early stages. The proposed approach is based on calculating the percentage of healthy, sick and dead regions of VF and OCT-GCL images. In order to carry out this calculation, we combined the thresholding methods with morphological operators and median filter to extract all regions. These algorithms developed were tested on a set of images of a local database composed of 58 OCT-GCL images and 21 VF images. The results obtained are satisfactory and confirmed by experts in ophthalmology.
关键词： Optical coherence tomography of ganglion cell layer,Segmentation,Visual field,Glaucoma,Characterization
更新于2025-09-23 15:22:29
-
[IEEE 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) - Guangzhou, China (2018.10.8-2018.10.12)] 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) - Automated Segmentation of Esophagus Layers from OCT Images Using Fast Marching Method
摘要： Thickness of the esophagus is an important diagnostic marker for many esophagus diseases. While labeling boundaries by manual to compute each layer’s average thickness is time-consuming and subjective. In this paper, we present a new fully automatic algorithm which includes Fast Marching Method (FMM) and Fourth-Order Runge-Kutta method (RK4) to identify five esophagus layers on optical coherence tomography (OCT) images. FMM is used to calculate the weighted geodesic distance. In particular, the velocity function involved in this method combines vertical gradient, horizontal gradient and curvature so that it not only can divide flat borders but also irregular borders. RK4 is used to find the shortest path which is the boundary to be segmented. The experimental comparison between automatic and manual is performed on 400 healthy guinea pig esophagus OCT images and the mean absolute error thickness difference between them is less than 6 pixels while the value can reach to 9.41 pixels at most between two observers.
关键词： Runge-Kutta method,Optical Coherence Tomography,Fast Marching Method,Image processing,Esophagus Layer Segmentation
更新于2025-09-23 15:22:29
-
Saliency Detection via Multi-Scale Global Cues
摘要： The saliency detection technologies are very useful to analyze and extract important information from given multimedia data, and have already been extensively used in many multimedia applications. Past studies have revealed that utilizing the global cues is effective in saliency detection. Nevertheless, most of prior works mainly considered the single-scale segmentation when the global cues are employed. In this paper, we attempt to incorporate the multi-scale global cues for saliency detection problem. Achieving this proposal is interesting and also challenging (e.g., how to obtain appropriate foreground and background seeds effectively? how to merge rough saliency results into the final saliency map efficiently?). To alleviate the challenges, we present a three-phase solution that integrates several targeted strategies: (i) a self-adaptive strategy for obtaining appropriate filter parameters; (ii) a cross-validation scheme for selecting appropriate background and foreground seeds; and (iii) a weight-based approach for merging the rough saliency maps. Our solution is easy-to-understand and implement, but without loss of effectiveness. Extensive experimental results based on benchmark datasets demonstrate the feasibility and competitiveness of our proposed solution.
关键词： image smoothing,global prior,segmentation,Saliency region
更新于2025-09-23 15:22:29
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens, Greece (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - Adversarial Domain Adaptation with a Domain Similarity Discriminator for Semantic Segmentation of Urban Areas
摘要： Existing semantic segmentation models of urban areas have shown to perform well in a supervised setting. However, collecting lots of annotated images from each city to train such models is time-consuming or difficult. In addition, when transferring the segmentation model from the trained city (source domain) to an unseen city (target domain), the performance will largely degrade due to the domain shift. For this reason, we propose a domain adaptation method with a domain similarity discriminator to eliminate such domain shift in the framework of adversarial learning. Contrary to the single-input adversarial network, our domain similarity discriminator, which consists of a Siamese network, is able to measure the similarity of the pairwise-input data. In this way, we can use more information about the pairwise-input to measure the similarity between different distributions so as to address the problem of domain shift. Experimental results demonstrate that our approach outperforms the competing methods on three different cities.
关键词： domain adaptation,urban areas,semantic segmentation,domain shift,Siamese network
更新于2025-09-23 15:22:29
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens, Greece (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - Dense Deconvolutional Network for Semantic Segmentation
摘要： Recently, exploring multiple feature maps from different layers in fully convolutional networks (FCNs) has gained substantial attention to capture context information for semantic segmentation. This paper presents a novel encoder-decoder architecture, called dense deconvolutional network (DDN), for semantic segmentation, where the feature maps of deeper convolutional layers are densely upsampled for the shallow deconvolutional layers. The proposed DDN is trainable end-to-end, and allows us to fully investigate multiple scale context cues embedded in images. The experimental results show that our DDN outperforms previous FCNs and encoder-decoder networks (EDNs) on PASCAL VOC 2012 dataset.
关键词： FCNs,EDNs,Semantic Segmentation,Dense Deconvolutional Network
更新于2025-09-23 15:22:29
-
[IEEE 2018 IEEE International Conference on Intelligent Transportation Systems (ITSC) - Maui, HI, USA (2018.11.4-2018.11.7)] 2018 21st International Conference on Intelligent Transportation Systems (ITSC) - Vehicle Detection and Localization using 3D LIDAR Point Cloud and Image Semantic Segmentation
摘要： This paper presents a real-time approach to detect and localize surrounding vehicles in urban driving scenes. We propose a multimodal fusion framework that processes both 3D LIDAR point cloud and RGB image to obtain robust vehicle position and size in a Bird's Eye View (BEV). Semantic segmentation from RGB images is obtained using our efficient Convolutional Neural Network (CNN) architecture called ERFNet. Our proposal takes advantage of accurate depth information provided by LIDAR and detailed semantic information processed from a camera. The method has been tested using the KITTI object detection benchmark. Experiments show that our approach outperforms or is on par with other state-of-the-art proposals but our CNN was trained in another dataset, showing a good generalization capability to any domain, a key point for autonomous driving.
关键词： localization,ERFNet,image semantic segmentation,KITTI,autonomous driving,vehicle detection,CNN,point cloud,multimodal fusion,3D LIDAR
更新于2025-09-23 15:22:29
-
[ACM Press SIGGRAPH Asia 2017 Posters - Bangkok, Thailand (2017.11.27-2017.11.30)] SIGGRAPH Asia 2017 Posters on - SA '17 - 4D computed tomography measurement for growing plant animation
摘要： Detailed observation of plant growth is essential for botanical analysis and realistic animation design. This study introduces spatial-temporal measurement techniques for a growing plant using X-ray Computed Tomography (CT). We scanned a target plant using CT over the course of couple of days with fixed time intervals to obtain four-dimensional (4D) volumetric images. We present a technique to segment the obtained 4D-CT images semi-automatically. We provide a 4D-CT measurement of budding bean sprouts to illustrate the feasibility of it.
关键词： Plant Animation,4D Measurement,Segmentation,X-ray CT
更新于2025-09-23 15:22:29
-
Automatic Pathological Lung Segmentation in Low-dose CT Image using Eigenspace Sparse Shape Composition
摘要： Segmentation of lungs with severe pathology is a nontrivial problem in clinical application. Due to complex structures, pathological changes, individual differences and low image quality, accurate lung segmentation in clinical 3D CT images is still a challenging task. To overcome these problems, a novel dictionary-based approach is introduced to automatically segment pathological lungs in 3D low-dose CT images. Sparse shape composition is integrated with eigenvector space shape prior model, called eigenspace sparse shape composition, to reduce local shape reconstruction error caused by weak and misleading appearance prior information. To initialize the shape model, a landmark recognition method based on discriminative appearance dictionary is introduced to handle lesions and local details. Furthermore, a new vertex search strategy based on gradient vector flow field is also proposed to drive shape deformation to target boundary. The proposed algorithm is tested on 78 3D low-dose CT images with lung tumors. Compared to state-of-the-art methods, the proposed approach can robustly and accurately detect pathological lung surface.
关键词： gradient vector flow,Pathological lung segmentation,discriminative appearance dictionary,eigenspace sparse shape composition
更新于2025-09-23 15:22:29