Since it seems like the poster works for the company: the ads on this article are all semi-nsfw for me, more-so than the average GAd I see (is that something you can tune?). I reported them to Google but that might be a black hole.<p>Additionally, why in the world does a technical blog for a for-profit company need ads?
It looks promising.<p>I like the way each algorithm is explained with minimal example and demo images.<p>I'm looking for such a lightweight vision lib to embed simple image manipulation programs on tiny ESP32-CAM boards.<p>OpenCV seems too heavy to integrate on such small devices.<p>So far, I've been able to develop simple image processing programs with CImg[1] (simple filtering, image cropping, adding text).<p>I'll try to spend some time exploring this SOD lib, thank you for sharing.<p>[1] <a href="https://cimg.eu/" rel="nofollow">https://cimg.eu/</a>
Visit the url below for much of the same content but without ads.
<a href="https://sod.pixlab.io/" rel="nofollow">https://sod.pixlab.io/</a>
Note this is GPL which will make it a harder sell vs OpenCV in many cases. I'd guess they're looking to use the non-GPL for $$ model.<p>Nevertheless, it looks interesting. It looks really nice for embedded where you don't always have the GPU for compute.
Is there anything substantially different here from the way any modern post-processing shader works right now? For example: <a href="https://threejs.org/examples/webgl_postprocessing_rgb_halftone.html" rel="nofollow">https://threejs.org/examples/webgl_postprocessing_rgb_halfto...</a>
How does it compare to dlib [1], ccv [2] etc.?<p>[1] <a href="https://github.com/davisking/dlib">https://github.com/davisking/dlib</a><p>[2] <a href="https://github.com/liuliu/ccv">https://github.com/liuliu/ccv</a>