Digital Measurement System for Ship Draft Survey

2013 ◽  
Vol 333-335 ◽  
pp. 312-316 ◽  
Author(s):  
Hong Wei Gu ◽  
Wang Zhang ◽  
Wen Hai Xu ◽  
Ying Li
Keyword(s):  
Measurement System ◽  
Color Image ◽  
Image Acquisition ◽  
Visual Observation ◽  
Mean Value ◽  
Color Image Segmentation ◽  
Digital Measurement ◽  
High Definition ◽  
The Mean ◽  
Image Segmentation Algorithm

In order to overcome the limits of visual observation widely used in ship draft survey, a digital measurement system based on image acquisition and processing is presented. The images of drafts are obtained by an image acquisition unit containing a high definition camera extended to the shipboard from the deck. The draft marks are digitalized in use of a multilayer neural network, and the draft lines are detected by means of color image segmentation algorithm. Finally, the ship's trim are identified based on the relative position of the draft line located at the digitalized draft marks, where the artificial influences are eliminated. The influences of wave may also be reduced by calculating the mean value of the ship's trim obtained from different images.

Synchronous Issue of IR-Based Large Scale Volume Localization System

2014 ◽  
Vol 496-500 ◽  
pp. 1643-1647
Author(s):  
Ying Feng Wu ◽  
Gang Yan Li
Keyword(s):  
Standard Deviation ◽  
Mobile Robot ◽  
Large Volume ◽  
Large Scale ◽  
Image Acquisition ◽  
Mean Value ◽  
Network Synchronization ◽  
Localization System ◽  
Localization Performance ◽  
The Mean

IR-based large scale volume localization system (LSVLS) can localize the mobile robot working in large volume, which is constituted referring to the MSCMS-II. Hundreds cameras in LSVLS must be connected to the control station (PC) through network. Synchronization of cameras which are mounted on different control stations is significant, because the image acquisition of the target must be synchronous to ensure that the target is localized precisely. Software synchronization method is adopted to ensure the synchronization of camera. The mean value of standard deviation of eight cameras mounted on two workstations is 12.53ms, the localization performance of LSVLS is enhanced.

Color Image Segmentation Based on Self-Organizing Maps

2011 ◽  
Vol 214 ◽  
pp. 693-698
Author(s):  
Rui Geng
Keyword(s):  
Image Segmentation ◽  
Optimization Problems ◽  
Color Image ◽  
Optimal Solution ◽  
Implementation Process ◽  
Segmentation Algorithm ◽  
Color Image Segmentation ◽  
Combinatorial Optimization Problems ◽  
Bee Colony ◽  
Image Segmentation Algorithm

The colony intellectual behavior performed by many organisms in nature can solve various kinds of problems on scientific and technological research. Bees are a socialized insect colony, which perform different types of activities according to their different divisions of labor, and achieve information sharing and exchanges among the bee colony to find the optimal solution for problems. According to this characteristic, researchers have proposed the algorithm of bee colony for solving combinatorial optimization problems. In this paper, it will describe the implementation process of such an image segmentation algorithm, and the result shows that this method is a potential image segmentation algorithm.

Fuzzy Clustering Color Image Segmentation Algorithm Based on CPSO

2012 ◽  
Vol 461 ◽  
pp. 526-531
Author(s):  
Xiao Hong Zhang ◽  
Hong Mei Ning
Keyword(s):  
Image Segmentation ◽  
Color Image ◽  
Segmentation Algorithm ◽  
Color Image Segmentation ◽  
Local Optimum ◽  
Fcm Clustering ◽  
Clustering Center ◽  
Membership Matrix ◽  
Image Segmentation Algorithm

Fuzzy C-mean algorithm (FCM) has been well used in the field of color image segmentation. But it is sensitive to initial clustering center and membership matrix, and likely converges into the local minimum, which causes the quality of image segmentation lower. By use of the properties-ergodicity, randomicity of chaos, a new image segmentation algorithm is proposed, which combines the chaos particle swarm optimization (CPSO) and FCM clustering. Some experimental results are shown that this method not only has the ability to prevent the particles to convergence to local optimum, but also has faster convergence and higher accuracy for segmentation. Using the feature distance instead of Euclidian distance, robustness of this method is enhanced.

Sign in / Sign up

Export Citation Format

Share Document