{"id":129,"date":"2026-03-24T15:06:30","date_gmt":"2026-03-24T13:06:30","guid":{"rendered":"https:\/\/blogs.uef.fi\/photogrammetry\/?page_id=129"},"modified":"2026-03-24T15:08:32","modified_gmt":"2026-03-24T13:08:32","slug":"gaussian-splatting","status":"publish","type":"page","link":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/","title":{"rendered":"Gaussian Splatting"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">From Points to Radiance: The Evolution of Gaussian Splatting in Photogrammetry<\/h2>\n\n\n\n<p>For decades, the photogrammetric pipeline has remained relatively stable: feature extraction, matching, bundle adjustment, and dense reconstruction. However, as we push toward real-time immersion and digital twins, traditional meshes often fall short\u2014struggling with non-Lambertian surfaces, fine geometries like hair, and computational overhead.<\/p>\n\n\n\n<p>Enter <strong>3D Gaussian Splatting (3DGS)<\/strong>. This technique isn\u2019t just a new way to render; it\u2019s a fundamental shift in how we represent captured reality.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"640\" height=\"360\" src=\"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif\" alt=\"\" class=\"wp-image-131\" \/><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\">The Technical Core: Beyond the Point Cloud<\/h3>\n\n\n\n<p>Traditional photogrammetry yields a sparse point cloud. 3DGS takes those points and transforms them into a collection of millions of 3D Gaussians. Each Gaussian is defined by a 3D covariance matrix \u03a3 and a mean position (center) \u03bc.<\/p>\n\n\n\n<p>To keep the representation differentiable and physically meaningful, the covariance matrix is decomposed into a scaling matrix S and a rotation matrix R (represented by a quaternion q):<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"165\" height=\"55\" src=\"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/image.png\" alt=\"\" class=\"wp-image-130\" \/><\/figure>\n\n\n\n<p>Each &#8220;splat&#8221; also carries:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Opacity (\u03b1):<\/strong> Determining how transparent the Gaussian is.<\/li>\n\n\n\n<li><strong>Spherical Harmonics (SH):<\/strong> Capturing view-dependent color (the &#8220;sheen&#8221; on a car or the flicker of sunlight on water).<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\">The 3DGS Pipeline for Researchers<\/h3>\n\n\n\n<p>The magic of 3DGS is that it leverages the best of photogrammetry to skip the &#8220;cold start&#8221; problem faced by many neural radiance fields.<\/p>\n\n\n\n<ol start=\"1\" class=\"wp-block-list\">\n<li><strong>SfM Initialization:<\/strong> You begin with standard photogrammetric software (like COLMAP) to estimate camera poses and generate a sparse point cloud.<\/li>\n\n\n\n<li><strong>Differentiable Rendering:<\/strong> Unlike NeRFs, which sample points along a ray (a costly process), 3DGS projects these 3D Gaussians into 2D &#8220;tiles&#8221; on the image plane.<\/li>\n\n\n\n<li><strong>Adaptive Density Control:<\/strong> During optimization, the system performs &#8220;cloning&#8221; and &#8220;splitting.&#8221; If a Gaussian has a high gradient, it\u2019s either split into two smaller ones (if the area is over-reconstructed) or moved (if it\u2019s under-reconstructed).<\/li>\n\n\n\n<li><strong>Stochastic Gradient Descent:<\/strong> The model minimizes the difference between the rendered splats and the original high-resolution photographs.<\/li>\n<\/ol>\n\n\n\n<h3 class=\"wp-block-heading\">Why This Matters for the Digital Twin Era<\/h3>\n\n\n\n<p>You likely see the trade-offs in current workflows. 3DGS offers three distinct advantages:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Training Speed:<\/strong> While a high-quality NeRF might take hours or days to converge, a 3DGS scene can often be &#8220;trained&#8221; in 20\u201340 minutes on a single consumer GPU.<\/li>\n\n\n\n<li><strong>Real-Time Inference:<\/strong> Because the rendering is essentially specialized rasterization, we can achieve 100+ FPS at 1080p resolution.<\/li>\n\n\n\n<li><strong>Geometric Fidelity:<\/strong> By using the SfM point cloud as a prior, we ensure the spatial coordinates remain grounded in the physical reality of the sensors.<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\">Conclusion: A New Standard?<\/h3>\n\n\n\n<p>We are moving away from the era of &#8220;hollow&#8221; meshes draped in flat textures. We are entering an era of volumetric probability. For researchers in photogrammetry, 3DGS isn&#8217;t a replacement for our foundational principles\u2014it is the ultimate refinement of them.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>From Points to Radiance: The Evolution of Gaussian Splatting in Photogrammetry For decades, the photogrammetric pipeline has remained relatively stable: feature extraction, matching, bundle adjustment, and dense reconstruction. However, as we push toward real-time immersion and digital twins, traditional meshes often fall short\u2014struggling with non-Lambertian surfaces, fine geometries like hair, and computational overhead. Enter 3D [&hellip;]<\/p>\n","protected":false},"author":746,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"footnotes":""},"class_list":["post-129","page","type-page","status-publish","hentry"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.3 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Gaussian Splatting - Learn Photogrammetry<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Gaussian Splatting - Learn Photogrammetry\" \/>\n<meta property=\"og:description\" content=\"From Points to Radiance: The Evolution of Gaussian Splatting in Photogrammetry For decades, the photogrammetric pipeline has remained relatively stable: feature extraction, matching, bundle adjustment, and dense reconstruction. However, as we push toward real-time immersion and digital twins, traditional meshes often fall short\u2014struggling with non-Lambertian surfaces, fine geometries like hair, and computational overhead. Enter 3D [&hellip;]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/\" \/>\n<meta property=\"og:site_name\" content=\"Learn Photogrammetry\" \/>\n<meta property=\"article:modified_time\" content=\"2026-03-24T13:08:32+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif\" \/>\n\t<meta property=\"og:image:width\" content=\"640\" \/>\n\t<meta property=\"og:image:height\" content=\"360\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/gif\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"3 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/\",\"url\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/\",\"name\":\"Gaussian Splatting - Learn Photogrammetry\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/wp-content\\\/uploads\\\/sites\\\/231\\\/2026\\\/03\\\/ezgif-5ed09c2ce7f92964.gif\",\"datePublished\":\"2026-03-24T13:06:30+00:00\",\"dateModified\":\"2026-03-24T13:08:32+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/#primaryimage\",\"url\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/wp-content\\\/uploads\\\/sites\\\/231\\\/2026\\\/03\\\/ezgif-5ed09c2ce7f92964.gif\",\"contentUrl\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/wp-content\\\/uploads\\\/sites\\\/231\\\/2026\\\/03\\\/ezgif-5ed09c2ce7f92964.gif\",\"width\":640,\"height\":360},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/gaussian-splatting\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Gaussian Splatting\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/#website\",\"url\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/\",\"name\":\"Learn Photogrammetry\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/blogs.uef.fi\\\/photogrammetry\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Gaussian Splatting - Learn Photogrammetry","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/","og_locale":"en_US","og_type":"article","og_title":"Gaussian Splatting - Learn Photogrammetry","og_description":"From Points to Radiance: The Evolution of Gaussian Splatting in Photogrammetry For decades, the photogrammetric pipeline has remained relatively stable: feature extraction, matching, bundle adjustment, and dense reconstruction. However, as we push toward real-time immersion and digital twins, traditional meshes often fall short\u2014struggling with non-Lambertian surfaces, fine geometries like hair, and computational overhead. Enter 3D [&hellip;]","og_url":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/","og_site_name":"Learn Photogrammetry","article_modified_time":"2026-03-24T13:08:32+00:00","og_image":[{"url":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif","width":640,"height":360,"type":"image\/gif"}],"twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"3 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/","url":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/","name":"Gaussian Splatting - Learn Photogrammetry","isPartOf":{"@id":"https:\/\/blogs.uef.fi\/photogrammetry\/#website"},"primaryImageOfPage":{"@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/#primaryimage"},"image":{"@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/#primaryimage"},"thumbnailUrl":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif","datePublished":"2026-03-24T13:06:30+00:00","dateModified":"2026-03-24T13:08:32+00:00","breadcrumb":{"@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/#primaryimage","url":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif","contentUrl":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-content\/uploads\/sites\/231\/2026\/03\/ezgif-5ed09c2ce7f92964.gif","width":640,"height":360},{"@type":"BreadcrumbList","@id":"https:\/\/blogs.uef.fi\/photogrammetry\/gaussian-splatting\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/blogs.uef.fi\/photogrammetry\/"},{"@type":"ListItem","position":2,"name":"Gaussian Splatting"}]},{"@type":"WebSite","@id":"https:\/\/blogs.uef.fi\/photogrammetry\/#website","url":"https:\/\/blogs.uef.fi\/photogrammetry\/","name":"Learn Photogrammetry","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/blogs.uef.fi\/photogrammetry\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"_links":{"self":[{"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/pages\/129","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/users\/746"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/comments?post=129"}],"version-history":[{"count":2,"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/pages\/129\/revisions"}],"predecessor-version":[{"id":134,"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/pages\/129\/revisions\/134"}],"wp:attachment":[{"href":"https:\/\/blogs.uef.fi\/photogrammetry\/wp-json\/wp\/v2\/media?parent=129"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}