Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
An approach to a pseudo real-time image processing engine for hyperspectral imaging

An approach to a pseudo real-time image processing engine for hyperspectral imaging

Ratings:

1.0

(1)
|Views: 197 |Likes:
Published by ijcsis
Hyperspectral imaging provides an alternative way of increasing the accuracy by adding another dimension: the wavelength. Recently, hyperspectral imaging is also finding its way into many more applications, ranging from medical imaging in endoscopy for cancer detection to quality control in the sorting of fruit and vegetables. But effective use of hyperspectral imaging requires an understanding of the nature and limitations of the data and of various strategies for processing and interpreting it. Also, the breakthrough of this technology is limited by its cost, speed and complicated image interpretation. We have therefore initiated work on designing real-time hyperspectral image processing to tackle these problems by using a combination of smart system design, and pseudo-real time image processing software. The main focus of this paper is the development of a camera-based hyperspectral imaging system for stationary remote sensing applications. The system consists of a high performance digital CCD camera, an intelligent processing unit, an imaging spectrograph, an optional focal plane scanner and a laptop computer equipped with a frame grabbing card. In addition, special software has been developed to synchronize between the frame grabber (video capture card), and the digital camera with different image processing techniques for both digital and hyperspectral data.
Hyperspectral imaging provides an alternative way of increasing the accuracy by adding another dimension: the wavelength. Recently, hyperspectral imaging is also finding its way into many more applications, ranging from medical imaging in endoscopy for cancer detection to quality control in the sorting of fruit and vegetables. But effective use of hyperspectral imaging requires an understanding of the nature and limitations of the data and of various strategies for processing and interpreting it. Also, the breakthrough of this technology is limited by its cost, speed and complicated image interpretation. We have therefore initiated work on designing real-time hyperspectral image processing to tackle these problems by using a combination of smart system design, and pseudo-real time image processing software. The main focus of this paper is the development of a camera-based hyperspectral imaging system for stationary remote sensing applications. The system consists of a high performance digital CCD camera, an intelligent processing unit, an imaging spectrograph, an optional focal plane scanner and a laptop computer equipped with a frame grabbing card. In addition, special software has been developed to synchronize between the frame grabber (video capture card), and the digital camera with different image processing techniques for both digital and hyperspectral data.

More info:

Published by: ijcsis on Nov 02, 2010
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

08/21/2013

pdf

text

original

 
 1
An approach to a pseudo real-time image processing engine forhyperspectral imaging
Abstract
 
Hyperspectral imaging provides an alternative way of increasingthe accuracy by adding another dimension: thewavelength. Recently, hyperspectral imaging is also finding itsway into many more applications, ranging from medical imagingin endoscopy for cancer detection to quality control in the sortingof fruit and vegetables.
 
But effective use of hyperspectralimaging requires an understanding of the nature and limitationsof the data and of various strategies for processing andinterpreting it. Also, the breakthrough of this technology islimited by its cost, speed and complicated image interpretation.We have therefore initiated work on designing real-timehyperspectral image processing to tackle these problems by usinga combination of smart system design, and pseudo-real timeimage processing software.
 
The main focus of this paper is thedevelopment of a camera-based hyperspectral imaging system for stationary remote sensing applications. The system consists of ahigh performance digital CCD camera, an intelligent processingunit, an imaging spectrograph, an optional focal plane scanner and a laptop computer equipped with a frame grabbing card.
 
Inaddition, special software has been developed to synchronize between the frame grabber (video capture card), and the digitalcamera with different image processing techniques for bothdigital and hyperspectral data
.
 
 Keywords:
 
 Remote sensing, image processing, Real-Time, frame grabber, hyperspectral, Hardware/Software Design.
 
1. Introduction
Digital and Remote sensing image processing is nowadaysa mature research area. Use of hyperspectral remotesensing in both research and operational applications has been steadily increasing in the last decade. Hyperspectralimaging systems can capture imagery from tens tohundreds of narrow bands in the visible to infrared spectralregions. These systems offer new opportunities for better differentiation and estimation of biophysical attributes andhave the potential for identification of optimal bandsand/or band combinations for a variety of remote sensingapplications [1-3],[11]. Different remote sensingapplications have proven to be potential sources of reflectance data such as extraction of various vegetationspectral features. Satellite-based remote sensing provides aunique opportunity to obtain characteristics over largeareas, whereas airborne remote sensing provides remotelysensed data over the medium scale, such as farms andsmall watersheds [4]. However, these studies largelydepend on the availability of spectral images that areusually quite expensive and need to be acquired by professional image providers. Ground based hyperspectralimaging has been used as a cheap tool to acquire remotelysensed data from individual part of proposed area [4].In this paper, we propose an approach to pseudoreal-time image processing engine for hyperspectralimaging
 
to increase mission flexibility for environmental planning,
 
medical diagnostics, remote sensing, and naturalresources applications. All processes in theimplementation of hyperspectral imagery and remotesensing apply near real time image processing done at thespatial and numerical modeling laboratory (SNML) at theUniversity of Putra Malaysia. The main focus of thisresearch is the development of a camera-basedhyperspectral imaging system for stationary remotesensing applications.
 
Hyperspectral imaging provides analternative way of increasing the accuracy by addinganother dimension: the wavelength. Recently,hyperspectral imaging is also finding its way into manymore applications, ranging from medical imaging inendoscopy for cancer detection to quality control in thesorting of fruit and vegetables.The impetus in performing this research wasgiven by existing snags and problems faced by workers inthe field. So far, many of the image processing softwareavailable in the market do not process images in real time.The software has to download and read the images firstand then prepare image-processing functionalities on them.In this paper, we attempt to show that it is possible to have pseudo-real image processing. This means that processingis done on the fly: as soon as the camera captures theimage, the image processing algorithm comes into playimmediately in all embedded applications.
* Responsible author 
Sahar Sabbaghi Mahmouei*
Smart Technology and Robotics ProgrammeInstitute of Advanced Technology (ITMA),Universiti Putra Malaysia, Serdang, Malaysiasabbaghi.sahar@gmail.com
Prof.Dr.Shattri Mansor
Remote Sensing and GIS Programme,Department of Civil Engineering,Universiti Putra Malaysia, Serdang, Malaysiashattri@eng.upm.edu.my
Abed Abedniya
MBA Programme,Faculty of Management (FOM),Multimedia University, Malaysiaabed.abedniya@gmail.com
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010201http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
 2
 The Hyperspectral imaging system consists of four components:
 
A sensing component: a hyperspectral sensor (high
 
 performance digital CCD camera nowknown as the ImSpector manufactured bySPECIM systems) for acquiring data or images.
 
An optional focal plane scanner 
 
A video capture (frame grabber) card connectedto the CPU from the camera helps in data capture.
 
Acer Extensa Notebook 4630z, which is a 2.0GHz Intel notebook, computer manufactured byAcer Inc., has been used as the CPU on the sensor  part.The remainder of this article is structured asfollows. Section 2 presents essential characteristics andconcepts in the scope of the work. Section 3 presentsSystem Requirement
.
We will describe developingsoftware in section 4. Description of proposed method,design and a relative technique are discussed in section 5.Section 6 shows the experimental result and discussions.Section 7 presents the conclusion of this paper.
2. Concepts and characteristics
In order to draw a clear picture of fundamental conceptsand characteristics of hyperspectral imaging, it isimportant to recap some key concepts and definitionswhich are accepted by experts in this field.
 Real-time image processing:
 
operating systemsserve application requests nearly real-time. In the other word manipulation of live images, typically within 50 to100 milliseconds, so the human user perceives them asinstantaneous.
 Embedded systems:
An embedded system isa computer system designed to perform one or a fewdedicated functions often with real-timecomputing constraints. It is
embedded 
as part of acomplete device often including hardware and mechanical parts.
 Engine:
 
The image processing engine, or image processor, is an important component of a digitalcamera and plays a vital role in creating the digital image.The image processing engine comprises a combination of hardware processors and software algorithms. The image processor gathers the luminance and chrominanceinformation from the individual pixels and uses it tocompute/interpolate the correct color and brightness valuesfor each pixel.
Pushbroom:
 
In remote sensing, an imaging deviceconsisting of a linear array of sensors (CCD camera)which is swept across the area of observation. Pushbroomsystems allow greater resolution of data to be assimilatedthan do line scanner systems.
3.
 
Real-Time Hyperspectral Imaging SystemRequirement
 
3.1. The sensor 
The hyperspectral sensor used in this study was a ground based user-friendly line sensor ImSpector (V10) See(Fig.1). The new ImSpector Fast10 is a high intensityimaging spectrograph, and makes spectral imaging possible at hundreds and even up to 1500 images per second. ImSpector Fast10 imaging spectrograph provides[5]:
 
high light throughput
 
superior image quality
 
good spectral resolution of 15 nm
 
full VNIR spectrum of 400 - 1000 nm over anarrow dimension, allowing short read out times
 
maximum light intensity on the camera pixels,allowing short integration times
 
high speed acquisition in many low cost industrialCCD and CMOS camerasThe ImSpector imaging spectrograph is acomponent that can be combined with a broad range of monochrome matrix cameras to form a spectral imagingdevice. Equipping the instrument with an objective lenscoupled with a monochrome area camera, convertsImSpector to a spectral line imaging camera. Operation is based on the direct sight imaging spectrograph technologyof the Spectral Imaging Ltd. (SPECIM), Oulu, Finland [6].ImSpector captures a line image of a target and disperseslight from each line image pixel to spectrum. Each spectralimage then contains line pixels in the spatial axis andspectral pixels in the spectral axis (Fig. 2) [4]. It is possibleto acquire full spectral information for each line imageacquired from the target. Since ImSpector capturessequential images of the moving target (or the sensor itself moves), a 2D spectral image can be formed. Thistechnology allows diverse opportunities to analyze thetarget accurately based on its spectral features.
Fig. 1– hyperspectral sensor (ImSpector V10).
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010202http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
 3
 
3.1.1 Advantages of hyperspectral imaging system
 
Hyperspectral imaging is extremely advantageous in termsof its data, presenting the information in the spatialdirection which is useful for extracting information withless loss of data. Some advantages of hyperspectralimaging over conventional techniques such as: NIRS(Near-infrared spectroscopy), RGB, and hyperspectralimaging are shown in Table 1 [7, 8].
Feature RGBimagingNIRS MSI HSISpatialinformation
 
 
 
 
Spectralinformation
 
 
Limited
 
Multi-constituent
 Limited
 
Limited
 
Sensitivity tominorComponents
Limited
 
Table.1 Advantages of hyperspectral imaging systemFig. 2 – The operating principles of ImSpector.
3.2 An optional focal plane scanner 
The focal plane scanner performs line scanning across aninput imaging area within the focal plane of the front lensand the spectrograph disperses each line into a spectrumand projects a two-dimensional image profile (line image)onto the CCD surface. This configuration allows imageacquisition under stationary or laboratory settings [2], [10].
3.3 A video capture (frame grabber)
The FrameLink frame grabber is a TYPE II PC Card with both a Camera Link and Card bus interface. It provides theability to capture digital video data from a ‘baseconfiguration’ Camera Link interface and transfer that datato host memory via a Card bus (PCI) interface. The framelink is a professional state of the art PCMCIA card busdigital video capture card, allowing user to display,capture, store and preview mega pixel video image (up to16 mega pixels) on the notebook computer [9].
 
TheImperx FrameLink video capture card is as shown in(Fig.3) below.
Fig.3
The IMPERX FrameLink Fast CardBus video capture (framegrabber) card. This picture has been taken from the official website of Imperx Inc.
3.4 The computer system
The computer is an Intel Pentium III (800 MHz) processor  based system with 250 GB hard drive. The operatingsystem on the computer is Microsoft Windows XP. A PCIinterface board provided with the imaging system isinstalled in a master PCI slot in the computer. The utilitysoftware is installed in the computer for complete cameracontrol, image acquisition and applies image processingtechnique.
 
The Acer Notebook computer is as shown in(Fig. 4) below.
Fig.4
 –
Different views of the Acer Extensa 4630z Notebook computer. This has been used as the CPU on our hyperspectral imagingsystem. These pictures have been obtained from Acer Inc.
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 7, October 2010203http://sites.google.com/site/ijcsis/ISSN 1947-5500

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->