Object detection in hyperspectral imagery by using K-means clustering algorithm with pre-processing

Alam, Mohammad S.; Elbakary, M. I.; ASLAN, Murat

doi:10.1117/12.717926

Cited by 7 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The k-means algorithm belongs to cluster analysis, which is used to divide a set of objects into several groups according to the similarity of each other, where the similar objects constitute a group. The k-means algorithm has proved useful in clustering and segmentation of hyperspectral images [21][22][23].…”

Section: Image Repartitionmentioning

confidence: 99%

Rapid and Nondestructive Classification of Cantonese Sausage Degree Using Hyperspectral Images

Wang

2019

Applied Sciences

View full text Add to dashboard Cite

Hyperspectral images are widely used in the food industry as a fast and non-destructive analytical technique. Cantonese sausage has a long history and is a very old food production and meat preservation technology. According to the physical and chemical characteristics of the sausage, the Chinese business industry standard SB/t10003-92 divides the sausage into three grades, which are called excellent grade, level 1, and level 2. In this paper, k-means is adopted first to separate two parts of the meat adaptively to improve the discriminant rate. The hyperspectral information of the near-infrared band is extracted by successive projections algorithm (SPA). The multiple linear regression (MLR) and partial least squares regression (PLSR) algorithms are used to classify the sausage grade. The experimental results show that the lean meat and fat of the sausage have different characteristics in the near-infrared band, and the modeling results have higher accuracy and anti-interference after separating lean meat and fat meat. The best model of sausage classification is using SPA-MLR method to model the fat region of Cantonese sausage; the prediction accuracy of which is 100%. It was found that the modeling results of fat were better than lean meat in both PLSR and SPA-MLR, which indicated that there were obvious differences in fat composition among different grades of sausage, and the fat of sausage was more suitable for classification.

show abstract

Section: Image Repartitionmentioning

confidence: 99%

Rapid and Nondestructive Classification of Cantonese Sausage Degree Using Hyperspectral Images

Wang

2019

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…It depends on the number of pixels and the complexity of the data. The iteration process continues until the cluster centroids converge [8,9]. There are two methods to initialize the cluster centers [10,11].…”

Section: K-means Algorithmmentioning

confidence: 99%

Improved training for target detection using Fukunaga-Koontz transform and distance classifier correlation filter

2008

Self Cite

View full text Add to dashboard Cite

In a FLIR image sequence, a target may disappear permanently or may reappear after some frames and crucial information such as direction, position and size related to the target are lost. If the target reappears at a later frame, it may not be tracked again because the 3D orientation, size and location of the target might be changed. To obtain information about the target before disappearing and to detect the target after reappearing, distance classifier correlation filter (DCCF) is trained manualy by selecting a number of chips randomly. This paper introduces a novel idea to eliminates the manual intervention in training phase of DCCF. Instead of selecting the training chips manually and selecting the number of the training chips randomly, we adopted the K-means algorithm to cluster the training frames and based on the number of clusters we select the training chips such that a training chip for each cluster. To detect and track the target after reappearing in the field-ofview ,TBF and DCCF are employed. The contduced experiemnts using real FLIR sequences show results similar to the traditional agorithm but eleminating the manual intervention is the advantage of the proposed algorithm.

show abstract

“…Yüzlerce dalga boyuna bakıldığında, farklı materyaller ışığı farklı şekillerde yansıttıklarından, yansıma spektrasına bakarak hiperspektral görüntülerde her bir pikselde hangi materyalin olduğunu anlamak mümkündür. Şekil 1'de örneklenen bu önemli bilgi, son yıllarda yeryüzü şekillerinin ayrıştırılması, maden yataklarının tespiti, tarım alanlarının analizi, savunma sanayiinde hedef tespiti ve benzeri birçok alanda kullanılmaktadır [1]- [6].…”

Section: Introductionunclassified

Effect of LiDAR sensor on the success of shadow detection from hyperspectral data

Yüksel¹,

Boyacı²

2018

Pamukkale J Eng Sci

View full text Add to dashboard Cite

Object detection in hyperspectral imagery by using K-means clustering algorithm with pre-processing

Cited by 7 publications

References 0 publications

Rapid and Nondestructive Classification of Cantonese Sausage Degree Using Hyperspectral Images

Rapid and Nondestructive Classification of Cantonese Sausage Degree Using Hyperspectral Images

Improved training for target detection using Fukunaga-Koontz transform and distance classifier correlation filter

Effect of LiDAR sensor on the success of shadow detection from hyperspectral data

Contact Info

Product

Resources

About