Çalışkan, Akın
BATI, Emrecan
Koz, Alper
Alatan, Abdullah Aydın
Using the spectral signature of a target by means of matching the signature with the pixels of an acquired hyperspectral image has been proven as an effective way of classifying hyperspectral pixels in most of the proposed methods in hyperspectral image analysis. A disadvantage of these methods is however to use only the spectral characteristics of pixels for detection while ignoring the spatial relations between the neighbouring pixels. In this paper, we propose a hyperspectral target detection method which uses also the spatial neigboorhood information as well as the spectral characteristics of hyperspectral pixels. To this end, we first utilize superpixelization method [1] to describe the neigborhood relation between the hyperspectral pixels, which has been previously developed and proved to be better compared to a pioneer state-of-the-art superpixel algorithm, SLIC [2]. Second, we investigate the best representatives for superpixels among different alternatives, such as centroids, medoid and mean, and modify the well-known hyperspectral target detection algorithm using orthogonal subspace projection, DTDCA [3], appropriately for superpixels. The improvements of the proposed approach over DTDCA in terms of the detection and false detection rates are verified on real hyperspectral images taken from wheat and corn fields with a VNIR camera.


Süperpikseller ve İmza Tabanlı Yöntemler Kullanarak Hiperspektral Hedef Tespiti
Kütük, Mustafa; Alatan, Abdullah Aydın (2019-06-27)
Spectral signature based methods which form the mainstream in hyperspectral target detection can be classified mainly in three categories as the methods using background modeling, subspace projection based methods, and hybrid methods merging linear unmixing with background estimation. A common characteristic of all these methods is to classify each pixel of the hyperspectral image as a target or background while ignoring the spatial relations between neighbor pixels. Integration of contextual information de...
Hyperspectral Superpixel Extraction Using Boundary Updates Based on Optimal Spectral Similarity Metric
Çalışkan, Akın; Koz, Alper; Alatan, Abdullah Aydın (2015-07-31)
The high spectral resolution of hyperspectral images (HSI) requires a heavy processing load. Assigning each pixel to a group in the image, which is called superpixel, and processing the superpixels instead of the pixels is resorted as a means to overcome this challenge in the hyperspectral literature. In this paper, we propose an algorithm to segment a hyperspectral image into superpixels by means of iteratively updating the boundary pixels of superpixels. We first explore the optimal similarity metric for ...
Superpixel Based Unsupervised Change Detection of Manmade Targets on Satellite Images
Buzcu, Ilker; Alatan, Abdullah Aydın (2015-05-19)
In this paper, a novel solution to the problem of change detection in bitemporal satellite images is presented. The approach can be described as 1) Preprocessing, in which both images' luminance levels are approximated to the same distribution via the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm and noise is filtered out by the use of a Bilateral Noise Reduction (BNR) filter; 2) Extraction of a joint superpixel map of both images using a novel method; 3) Calculation of a variety of met...
Özdemir, Okan Bilge; Soydan, Hilal; Çetin, Yasemin; Duzgun, Sebnem (2016-07-15)
This paper presents a vegetation detection application with semi-supervised target detection using hyperspectral unmixing and segmentation algorithms. The method firstly compares the known target spectral signature from a generic source such as a spectral library with each pixel of hyperspectral data cube employing Spectral Angle Mapper (SAM) algorithm. The pixel(s) with the best match are assumed to be the most likely target vegetation locations. The regions around these potential target locations are furt...
Post processing for wavelet domain HMT image resolution enhancement
Temizel, Alptekin (2007-06-13)
Wavelet domain image resolution enhancement algorithms assume that the available image is the low-frequency subband of a higher resolution image and high-frequency subbands are not available. Then, these high-frequency coefficients are estimated and the higher resolution image is generated by application of inverse wavelet transform. Some of these techniques have used probabilistic methods and utilisation of HMT (Hidden Markov Tree) was shown to produce promising results. HMT based methods model the wavelet...
Citation Formats
A. Çalışkan, E. BATI, A. Koz, and A. A. Alatan, “SUPERPIXEL BASED HYPERSPECTRAL TARGET DETECTION,” 2016, Accessed: 00, 2020. [Online]. Available: