Infopost | 2019.05.23

Thumbnail selection algorithm areas of interest motorcycles

On a whim I decided to update my thumbnail algorithm. In short, the previous iteration would look for areas of sharp contrast from a few preconfigured areas of the image. Intuitively, this had the twofold benefit of finding high contrast areas (sometimes where the important stuff is) and finding sharp areas (as out-of-focus parts won't have sharp contrast). I decided to iterate on this and preserve the sharp contrast element of the equation, but added parameters to try to resolve some of the issues with the original algorithm.

Image histogram

Both new ingredients are based on creating a histogram of brightness values using luminance (greyscale that treats R/G/B as the human eye does).
Rather than use preconfigured sample locations, I wrote the interface to let the user define a number of circlular sampling areas spaced evenly across the image. Running the algorithm would give me:

( x,   y): [score] [values...]
(88,  76): [265]   midtone: 96 contrast: 96   sharpness: 73
(88, 152): [104]   midtone: 40 contrast: 40   sharpness: 24
(88, 228):  [88]   midtone: 48 contrast: 24   sharpness: 16
(88, 304): [102]   midtone: 64 contrast: 24   sharpness: 14
(88, 380):  [79]   midtone: 48 contrast: 24   sharpness:  7
(88, 456):  [86]   midtone: 48 contrast: 24   sharpness: 14

Taking the combined midtone/contrast/sharpness values, I normalized them using them to give the percentage thumbnailability compared to the rest of the image.

Thumbnail selection algorithm areas of interest dog weimaraner

It gets a little clearer and more aligned when I draw on the sampling circles.

Thumbnail selection algorithm areas of interest volleyball

Generating a thumbnail is fairly trivial from here, but still heuristic. I could take the point of highest interest and expand out if necessary, but that could fall into an adversarial case fairly easily. Rather, iterating over the entire grid is pretty straightforward, and I can take the maximum grouping that meets provided size constraints.

Thumbnail selection algorithm areas of interest Dying Light screenshot

And beyond the thumbnail application, the high/low interest region labeling could be used for cropping or culling datasets for other algorithms.

Thumbnail selection algorithm areas of interest Risk of Rain 2 screenshot

Video games break the heuristic, but not as easily as the last one. Text provides the sharp contrast that the code likes to hone in on. Creating a center left/right bias would help with this, but then I need to provide manual or automatic recognition that it's a screencap.

Thumbnail selection algorithm areas of interest rafting

Even photos don't always get the intended result.

Related - internal

Some posts from this site with similar content.



I've done a little more work with my graphics library, following a few threads:


No, I'm not nerding out about newest Persona game, I'm nerding out about site meta! (There's the thin connection that it's about people and in the video game when a character attacks they sometimes shout "persona" because it's a jrpg). Anyway, I wante...


A short story from graphics to parallelism to lambda to mapreduce to hadoop to Spliterator to ParallelStream.

Related - external

Risky click advisory: these links are produced algorithmically from a crawl of the subsurface web (and some select mainstream web). I haven't personally looked at them or checked them for quality, decency, or sanity. None of these links are promoted, sponsored, or affiliated with this site. For more information, see this post.


Making Blue Noise Point Sets With Sliced Optimal Transport « The blog at the bottom of the sea

This post explains how to use sliced optimal transport to make blue noise point sets. The plain, well commented C++ code that goes along with this post, which made the point sets and diagrams, is at This is an implementation and investigation of "Sliced Optimal Transport Sampling" by Paulin et al ( also have...

Classic Computer Science Problems in Python | Henrik Warne's blog

I really enjoyed Classic Computer Science Problems in Python by David Kopec. It covers many different problems I hadn't read detailed explanations of before. For example: neural networks, constraint-satisfaction problems, genetic algorithms and the minimax algorithm. Unlike many other books on algorithms and programming problems, this one builds up complete (but small) programs that are...

Dynamic Programming is not Black Magic Quentin Santos

Created 2024.05 from an index of 235,542 pages.