Measuring Overlap-Rate in Hierarchical Cluster Merging for Image Segmentation and Ship Detection

2007 ◽  
Author(s):  
Zhiling Hong ◽  
Qingshan Jiang ◽  
Heshan Guan ◽  
Fangfei Weng
Keyword(s):  
Image Segmentation ◽  
Hierarchical Cluster ◽  
Ship Detection ◽  
Cluster Merging ◽  
Overlap Rate

IMAGE THRESHOLDING BASED ON HIERARCHICAL CLUSTERING ANALYSIS AND PERCENTILE METHOD FOR TUNA IMAGE SEGMENTATION

2018 ◽  
Vol 2 (1) ◽  
Author(s):  
Alifia Puspaningrum ◽  
Nahya Nur ◽  
Ozzy Secio Riza ◽  
Agus Zainal Arifin
Keyword(s):  
Image Segmentation ◽  
Hierarchical Clustering ◽  
Gabor Filter ◽  
Hierarchical Cluster ◽  
Experimental Result ◽  
Main Process ◽  
Percentile Method ◽  
Cluster Analysis Method ◽  
2D Gabor Filter

Automatic classification of tuna image needs a good segmentation as a main process. Tuna image is taken with textural background and the tuna’s shadow behind the object. This paper proposed a new weighted thresholding method for tuna image segmentation which adapts hierarchical clustering analysisand percentile method. The proposed method considering all part of the image and the several part of the image. It will be used to estimate the object which the proportion has been known. To detect the edge of tuna images, 2D Gabor filter has been implemented to the image. The result image then threshold which the value has been calculated by using HCA and percentile method. The mathematical morphologies are applied into threshold image. In the experimental result, the proposed method can improve the accuracy value up to 20.04%, sensitivity value up to 29.94%, and specificity value up to 17,23% compared to HCA. The result shows that the proposed method cansegment tuna images well and more accurate than hierarchical cluster analysis method.

scholarly journals TOWARDS DISTILLATION OF DEEP NEURAL NETWORKS FOR SATELLITE ON-BOARD IMAGE SEGMENTATION

2020 ◽  
Vol XLIII-B2-2020 ◽  
pp. 1553-1559
Author(s):  
F. de Vieilleville ◽  
A. Lagrange ◽  
R. Ruiloba ◽  
S. May
Keyword(s):  
Neural Networks ◽  
Image Segmentation ◽  
Data Analysis ◽  
Deep Neural Networks ◽  
Distillation Method ◽  
Ship Detection ◽  
Promising Solution

Abstract. Cubesats platforms expansion increases the need to simplify payloads and to optimize downlink data capabilities. A promising solution is to enhance on-board software, in order to take early decisions, automatically. However, the most efficient methods for data analysis are generally large deep neural networks (DNN) oversized to be loaded and processed on limited hardware capacities of cubesats. To use them, we must reduce the size of DNN while accommodating efficiency in terms of both accuracy and inference cost. In this paper, we propose a distillation method which reduces image segmentation deep neural network’s size to fit into on board processors. This method is presented through a ship detection example comparing accuracy and inference costs for several networks.

scholarly journals A PREALIMINARY STUDY OF SHIP DETECTION FROM UAV IMAGES BASED ON COLOR SPACE CONVERSION AND IMAGE SEGMENTATION

2017 ◽  
Vol XLII-2/W6 ◽  
pp. 189-193 ◽  
Author(s):  
A. M. Klimkowska ◽  
I. Lee
Keyword(s):  
Image Segmentation ◽  
Fishery Management ◽  
Color Space ◽  
Traffic Monitoring ◽  
Color Spaces ◽  
Monitoring And Control ◽  
Ship Wake ◽  
Ship Detection ◽  
Ship Wakes ◽  
The Impact

Ship detection is an inherent process supporting tasks such as fishery management, ship search, marine traffic monitoring and control, and helps in the prevention of illegal activities. So far, sea and shore monitoring has been carried out by ship patrols and aircrafts along with sea vessel detection from data from space-borne platforms. Recently an increase interest in applying images delivered by UAV for marine application due to their advantages such as high spatial resolution, independence on time acquisition can be noticed. While investigating state of the art methods used for ship detection from different platforms using optical images, we found a significant problem with occurrence of a ship wake. This phenomena may prohibit correct detection of ship location and results in overestimating the ship size as the ship and its wake are often considered as being part of the same object in image or wakes are distinguished as a separate ship due to their possible similar brightness compared with sea vessel. In order to reduce the impact of ship wakes we investigated the behavior of images in different color spaces to provide data with little or almost no trace of ship wake. We took into consideration following color spaces: HSV, YCbCr, NTSC, XYZ and L*a*b and investigated each channel from new images. Finally we decided to use 2nd channel of L*a*b space where the ship wakes occurrence were significantly reduced. Object of interest were detected through the use of image segmentation. Applied method uses edge detection based on the gradient magnitude calculation. Afterwards several characteristics such as centroids, major and minor axis, size and orientation were calculated for later use to remove false positives and thus improve accuracy of the proposed method.

Sign in / Sign up

Export Citation Format

Share Document