- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE 2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC) - Chongqing (2018.6.27-2018.6.29)] 2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC) - Intelligent Defogging Method Based on Clustering and Dark Channel Prior
摘要： In view of the foggy images collected by outdoor vision system are blurred, an intelligent defogging method based on clustering and dark channel prior is proposed. This method improves the traditional K-means clustering algorithm, considering the correlation between the samples and the running time of the algorithm, using the improved K-means clustering algorithm to recognize the foggy images; The traditional defogging algorithm based on dark channel prior is enhanced from the angle of improving the adaptability and efficiency of the algorithm, as well as improving the defogging effect, the clearness of the foggy images is realized based on the enhanced defogging algorithm. The Simulation results show that the proposed method can automatically recognize and process the foggy images, the recognition accuracy of the foggy images is high and the defogging effect is good, which is beneficial to improve the reliability of the outdoor vision system.
关键词： dark channel prior,defogging,clustering,recognition,intelligent
更新于2025-09-10 09:29:36
-
[IEEE 2018 International CET Conference on Control, Communication, and Computing (IC4) - Thiruvananthapuram, India (2018.7.5-2018.7.7)] 2018 International CET Conference on Control, Communication, and Computing (IC4) - Computer Aided Detection of Demarcation Line/Ridges in the Retinal Fundus Images of Preterm Infants
摘要： Retinopathy of Prematurity (ROP) is a curable sight threatening disease which affects the visual system of premature infants, weighing less than 1750 grams and who are born before 31 weeks of gestation. The characteristic features of the disease include abnormal blood vessel growth and scar tissue formation in the retina which may lead to permanent blindness. The risk of ROP can be reduced if the disease is diagnosed in the early stage of its development. As per the recommendations of the International Committee for ROP classification, the detection of demarcation line/ridge have great prognostic significance as its presence indicates the beginning stage of the disease. In the reported work we develop a computer aided diagnostic system for the early detection of ROP, where the presence of ridge/demarcation line is identified in the fundus images of the preterm infants. The developed system uses image enhancement techniques followed by clustering operation. An image database with 33 retinal images of preterm infants are graded with the proposed system and the results obtained matches with that of the clinical expert annotation with a good accuracy.
关键词： demarcation line,Retinopathy of prematurity,clustering
更新于2025-09-10 09:29:36
-
Location Ambiguity Resolution and Tracking Method of Human Targets in Wireless Infrared Sensor Network
摘要： Human tracking has attracted extensive attention by using low-cost pyroelectric infrared sensor network in recent years. This paper presents a location ambiguity resolution and tracking method for human targets in wireless, distributed and binary infrared sensor network. The tracking system can detect the human targets in the detection space, and activate the sensor detection lines dynamically. A bearing-crossing location method is designed. The intersections of all activated detection lines are called primary measurement points for human location, and some of them are false measurement points. The ambiguity of this bearing-crossing location method is discussed and a two-level bearing-crossing algorithm is proposed based on quartic K-means clustering and joint cost function. For the first level, an anti-logic algorithm is designed to get the initial effective measurement points, then these points are assigned to different targets using K-means clustering. For the second level, the final effective points are obtained by using a special joint cost function, and they are assigned to different targets using K-means clustering once again to get the final locating results. The cost value is used as a weight to adjust the covariance parameter in Kalman filter for target tracking as well. The experimental results show that the average tracking error of human targets is less than 0.8 m in a 10 m×10 m space, which verify the proposed location ambiguity resolution and tracking method.
关键词： Wireless Infrared sensor network,Cost function,Multiple human tracking,Binary pyroelectric infrared sensor network,Location ambiguity,Bearing-crossing location,Quadratic K-means clustering
更新于2025-09-10 09:29:36
-
[IEEE 2018 20th International Conference on Transparent Optical Networks (ICTON) - Bucharest (2018.7.1-2018.7.5)] 2018 20th International Conference on Transparent Optical Networks (ICTON) - Examples of Machine Learning Algorithms for Optical Network Control and Management
摘要： Machine learning (ML) offers a great variety of algorithms that can be used in the context of optical networks. In particular, ML algorithms might be applied for classification and to detect patterns, among others. Both, can help to facilitate improving its performance, as well as to understand the behavior of optical networks. In this paper, we review two of these ML algorithms, one for classification and the other for clustering. Illustrative examples of the application of such supervised and unsupervised ML algorithms applied to optical networks are presented.
关键词： support vector machine,machine learning,data visualization,clustering
更新于2025-09-10 09:29:36
-
[IEEE 2018 International CET Conference on Control, Communication, and Computing (IC4) - Thiruvananthapuram, India (2018.7.5-2018.7.7)] 2018 International CET Conference on Control, Communication, and Computing (IC4) - Reversible and Secure False Color Based Image Privacy Protection with Color Palette Generation
摘要： In many areas, visual privacy needs the protection, especially in video surveillance. Video surveillance is considered as an important solution for the security and safety issues in public places, since it monitors behaviour, activities, or other changing information. Its usage is increased in today’s world which affects individual’s privacy. There comes the need of visual privacy protection in video surveillance. Local servers can not store and analyze this large amount of data from the surveillance camera. So usage of cloud servers can solve this problem. But there is a chance to acquire these data by unauthorized parties. Different schemes for data hiding are present, which reversibly encrypts this data. But such systems need to do the detection of sensitive regions either by using a computer vision module or manually which increases complexity and reduces reliability. So two fully reversible privacy protection schemes for images using false coloring can be used. First scheme is more generic i.e., it is ?exible with other privacy protection schemes and the next scheme is fully based on false colors. These schemes are independent of region-of-interest (ROI) detection and can be applied to the whole image. So the user can escape from the complex ROI detection. A color palette generation method is also added with these schemes which increases the security.
关键词： k-means clustering,scrambling,pixelation,ROI,blurring,privacy protection,false coloring
更新于2025-09-10 09:29:36
-
[IEEE 2018 Second International Conference on Inventive Communication and Computational Technologies (ICICCT) - Coimbatore (2018.4.20-2018.4.21)] 2018 Second International Conference on Inventive Communication and Computational Technologies (ICICCT) - Detection of Pests Using Color Based Image Segmentation
摘要： Large amounts of crops are destroyed every year due to pests. Pest detection and identification is needed to ensure good productivity in agricultural crops. Early detection of pests in images is very crucial for effective management of pest control. Therefore, identifying the pest in the image has been a challenging task. In this paper, we use colour based image segmentation method to efficiently detect the pest. The extensive simulation results on various pest images show that the proposed method outperforms the existing Otsu’s method and edge detection segmentation.
关键词： edge detection,otsu’s thresholding,clustering,color based segmentation,pre-processing
更新于2025-09-10 09:29:36
-
Community-scale multi-level post-hurricane damage assessment of residential buildings using multi-temporal airborne LiDAR data
摘要： Building damage assessment is a critical task following major hurricane events. Use of remotely sensed data to support building damage assessment is a logical choice considering the di?culty of gaining ground access to the impacted areas immediately after hurricane events. However, a remote sensing based damage assessment approach is often only capable of detecting severely damaged buildings. In this study, an airborne LiDAR based approach is proposed to assess multi-level hurricane damage at the community scale. In the proposed approach, building clusters are ?rst extracted using a density-based algorithm. A novel cluster matching algorithm is proposed to robustly match post-event and pre-event building clusters. Multiple features including roof area and volume, roof orientation, and roof shape are computed as building damage indicators. A hierarchical determination process is then employed to identify the extent of damage to each building object. The results of this study suggest that our proposed approach is capable of 1) recognizing building objects, 2) extracting damage features, and 3) characterizing the extent of damage to individual building properties.
关键词： Hurricane damage assessment,Point cloud processing,Geometric computing,Airborne LiDAR,Data clustering
更新于2025-09-10 09:29:36
-
An image segmentation method using logarithmic kbest gravitational search algorithm based superpixel clustering
摘要： Image segmentation partitions an image into coherent and non-overlapping regions. Due to variations of visual patterns in images, it is a challenging problem. This paper introduces a new superpixel-based clustering method to efficiently perform the image segmentation. In the proposed method, initially superpixels from an image are obtained. The superpixels are further clustered into the required number of regions by a newly proposed variant of gravitational search algorithm namely; logarithmic kbest gravitational search algorithm. Experiments are conducted on the Berkeley Segmentation Dataset and Benchmark (BSDS500). It is affirmed from both visual and numerical analyses that the proposed method is efficacious and accurate in segmenting an image than the other considered segmentation methods.
关键词： BSDS500,Gravitational search algorithm,Kmeans,Superpixel clustering
更新于2025-09-10 09:29:36
-
BRDF Analysis with Directional Statistics and Its Applications
摘要： Data-driven BRDF models using real material measurements have become increasingly prevalent due to the development of novel goniore?ectometers, but ef?cient use of these models in many graphical applications remains challenging due to the few functionalities the raw data could provide. To ameliorate this issue, we propose to analyze BRDFs using directional statistics for better handling and exploring measured materials, especially isotropic materials, with ef?cient computation and compact storage. We conduct a thorough statistical analysis on both analytical BRDF models and measured materials from the MERL database. We show that different aspects of visual appearance can be characterized by different spherical moments, from which several descriptive measures can be derived to further facilitate their usage. We demonstrate how these measures are best leveraged in some graphical applications including gamut mapping using a new BRDF similarity measure, BRDF or SVBRDF reconstruction based on material clustering, and importance sampling for measured materials based on fast extracted GGX distributions. We ?nally show the potential of our approach in the categorization of surface re?ectance types which is common for traditional photon mapping.
关键词： rendering,clustering,BRDF,directional statistics,importance sampling
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Tabular K-Means Clustering on Remote Sensing Images
摘要： This research develops a Tabular K-means clustering approach that derives a discriminant look-up table (LUT) from the Voronoi diagram of initial K peaks automatically selected from the scatter plot of the top two principal components of the input images. Numerical experiments in clustering 7-band Landsat TM images into specified number of spectral clusters are illustrated for the advantages in convergence and computational efficiency of the proposed tabular approach against traditional approach in K-means clustering.
关键词： Voronoi diagram,peak detection,principal component transformation,K-means,clustering
更新于2025-09-10 09:29:36