<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Ongun on Medium]]></title>
        <description><![CDATA[Stories by Ongun on Medium]]></description>
        <link>https://medium.com/@ongun_32587?source=rss-18ebcbaa86a3------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*YjBu32ofdtvZrZ_qy7eRMw.png</url>
            <title>Stories by Ongun on Medium</title>
            <link>https://medium.com/@ongun_32587?source=rss-18ebcbaa86a3------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 23:52:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ongun_32587/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Seeing the unseen… What is Hyperspectral Imaging?]]></title>
            <link>https://medium.com/@ongun_32587/seeing-the-unseen-what-is-hyperspectral-imaging-d8c56212dd8a?source=rss-18ebcbaa86a3------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8c56212dd8a</guid>
            <dc:creator><![CDATA[Ongun]]></dc:creator>
            <pubDate>Mon, 13 Sep 2021 15:46:38 GMT</pubDate>
            <atom:updated>2022-10-04T11:17:44.980Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Image from — https://livingoptics.co/" src="https://cdn-images-1.medium.com/max/1024/1*Rm-9DnbGlHdTrIl5cdifdQ.gif" /><figcaption>GIF from <a href="https://livingoptics.co/">livingoptics.co</a>, an Oxford-based HSI startup</figcaption></figure><p><strong>Hyperspectral Imaging</strong> technology has been around since the mid-1980s and was derived from NASA’s Airborne Imaging Spectrometer (AIS) and Airborne Visible / InfraRed Imaging Spectrometer (AVIRIS) programmes. With recent developments, it presents the <strong>next stage of computer vision</strong>.</p><h3><strong>How does it work?</strong></h3><p>Essentially, HSI allows us to identify the chemical constitution of composite matter via an image. We call the study of the interaction between electromagnetic radiation and matter ‘<strong>spectroscopy</strong>’. As each chemical matter has its own unique indication called a ‘<strong>spectral signature’</strong> we can identify what forms of matter there are in each pixel of an image as shown below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0mxRjFHjq1tcwcx0a2_XCg.gif" /><figcaption>GIF from <a href="https://livingoptics.co/">livingoptics.co</a></figcaption></figure><p>The current pervasive form of RGB imaging assigns each pixel as either Red, Green or Blue, whereas HSI constructs a much larger spectrum of radiation (usually visible light to infrared) in each pixel meaning it contains over 10x more data. RGB imaging, therefore, is an inferior form of spectroscopy in most aspects, as it inherently has fewer data points to enable analysis. Though this technology is still unviable for commercial cameras, it has a plethora of use cases in many sectors.</p><p>For example, NASA&#39;s Hubble satellite which only captures visible light will soon be replaced by its successor, the ‘James Webb’ satellite, which will increase its perceivable wavelength to infrared and allow both hyper and multispectral analysis of the many astrophysical wonders of the galaxy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*HtN7sh0CovX6XA4n3I8tUg.gif" /></figure><h3>The next generation of computer vision?</h3><p>Previously, among other things, this form of spectroscopy has been used to find out what makes up our sun and other stars, but the applications are slowly being adapted for a wider set of uses. HSI cameras have become much cheaper to produce in recent years, with startups such as <a href="https://livingoptics.co/index.html"><strong><em>Living Optics</em></strong></a> claiming to be ready for mass production of small handheld devices. Thanks to <a href="https://en.wikipedia.org/wiki/Moore%27s_law"><strong><em>Moore’s law</em></strong></a>, we are also surpassing the processing bottleneck previously capping our ability to analyse these incredibly large sets of data. Developments in AI &amp; ML are also key to why HSI presents the biggest opportunity in computer vision in the years to come.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*M50f-UcqncGWvqc4D68D6g.png" /><figcaption>Click to enlarge</figcaption></figure><h3><strong>What are these applications?</strong></h3><p>The applications are endless for HSI. This technology has an overabundance of use cases across many sectors.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/617/1*dahqDj_1F-oEEEts401VhA.png" /><figcaption>Terrain mapping using HSI</figcaption></figure><p><em>For example, </em><a href="https://www.pixxel.space/"><em>Pixxel</em></a><em> is utilising this revolutionary tech to launch satellites fitted with HSI cameras. This enables them to map out terrains and geological hotspots for minerals, oil spills/gas seepages as well as plan urban infrastructures and map road networks.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CzqYwWYJQTmOR6ADhcZooQ.jpeg" /></figure><p><em>In Agg/FoodTech HSI is being used by many startups to see if crops are ripe/spoilt, monitor soil health and detect pest infestations.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/357/1*FxlMQatO_CP9S5iJvWpstA.jpeg" /></figure><p><em>In healthcare, this innovative technology is being used as a non-invasive form of diagnosis by startups such as </em><a href="http://www.bohndieklab.org/"><em>VisionLab</em></a><em>. Tumorous cells have a slight variation in their spectral signatures, so it may soon be possible to detect them via HSI</em>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dxkqhl7jGEt_Qa2sSFfk4A.png" /></figure><p><em>Use cases also expand into the pharmaceutical industry to detect the quality of the products as well as spot defects.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/734/1*vi8HTeQ-ct3myu1SEt4JDw.jpeg" /></figure><p><em>Some other uses are within archaeology, where it is used to extract additional information from artefacts as well as to scan the terrain to find new sites. Lastly, it also has military applications as it can be used to survey geographies and detect mines / camouflaged vehicles.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Wlg-Ft8Yea2fVJBeyuOLCw.png" /></figure><p><em>It is worth noting that this solution is not perfect for every problem. In recycling, for example, infrared sensor-based solutions are slowly losing the battle to RGB imaging-based startups such as Recycleye and Greyparrot which use AI to automate waste recognition systems. Said solutions are heavily dependent on building and training neural nets to detect recyclable waste. Currently, using HSI cameras is admittedly a much more expensive kit of hardware and RGB cameras in conjunction with ML seem to do the job for much cheaper.</em></p><h3><strong>What&#39;s next?</strong></h3><p>The possibilities are endless, as HSI can scale horizontally into multiple sectors. Although HSI is not the best solution in every case, its unique strengths in many applications present a compelling investment opportunity.</p><p>Unfortunately, VC funding in this space has been severely lacking in recent years. Lightspeed Ventures and Techstars are among the first to back this promising DeepTech and many are hoped to follow as more uses are commercialised.</p><p>If Theranos reached the heights it did with no functional technology in its Edison machine, what is the limit for HSI powered solutions in the years to come and who will lead the vanguard?</p><p>I certainly am looking forward to finding out.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8c56212dd8a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>