procerand.bsky.social
Alaskan Game Developer and Software Engineer. Currently working on an automated gigapixel scanner, a game, Lego sorter, alongside various 3D printing projects. (he/him)
https://linktr.ee/procerand
pfp made by @aroma031416.bsky.social
457 posts
68 followers
63 following
Regular Contributor
Active Commenter
comment in response to
post
That's cause like 85% of wikis use fandom which is infested with ads
comment in response to
post
Problem is, I'll likely be going from Alaska to Seattle so getting rid of the storage containers makes sense, especially since I know people in Alaska that are willing to mail my remaining stuff to me after I move and get settled down.
comment in response to
post
Yeah, I'm going to be running into that issue soon as well. I'm going to be moving in the next 6 months and I'm not entirely too sure if I'll end up having to get rid of a bunch of my storage containers to fit into a new space or if I'll have to get rid of all of them in order to move to a new state
comment in response to
post
I'm thankfully starting to reach the point where it's possible for me to start doing projects without buying components due to having them on hand since generally buying stuff online I get like a pack of 5 or 10 as its ends up being cheaper in the long run that way.
comment in response to
post
Totally up stuff I bought online just for the belt, I think I've spent ~$70 total on the belt including the motor driver required to actually drive it, but excluding the power supply since I have other 12v motors required for this project.
comment in response to
post
If the lego belt ends up wearing out, I'll likely switch to a cloth belt as other people have had success using cloth to make a belt before. Either way, the cost of building it myself with materials I own + some stuff bought online comes out to be much lower than buying a commercially available belt
comment in response to
post
And considering the complexity of a conveyor belt (dead simple) it felt kinda like a no brained to just make my own. The main issue is the belt material as I wasn't sure what to use. I originally tried a paper belt, but that kept tearing itself apart so then I moved onto using the lego belt.
comment in response to
post
If I recall correctly, the price to buy a conveyor belt with the specs I wanted was somewhere between $700 and $1000 which also just ever so happened to be approximately the amount of money I'm willing to spend on this project.
comment in response to
post
Yeah, being consistent with how the images are taken ends up solving a lot of issues. While this wouldn't work for images taken while holding the camera, it works perfectly when they're taken by a machine and you can use it's movement information to assist the image stitching process.
comment in response to
post
Yeah, initially I tried finding an off the shelf solution, but I couldn't quite find anything that matched the width + length I needed at a low cost. I might end up redoing the end piece again at a later date in order to free up the last 3 inches or so, but for now it's good enough.
comment in response to
post
Clarification: Same camera offset. The actual offset of the images themselves can vary a bit as long as they have moved the same distance each time.
comment in response to
post
This allows the program to essentially force the image into an approximate "proper" place and then rerun the image stitching algorithm for that section in order to center it correctly. However, this only works if the same offset is used for all images when taking them (which is extremely common).
comment in response to
post
For Forge, I've been working on doing a complete GUI rewrite making it easier to expand and use. As for the image stitching algorithm, I've been able to prevent large errors from occurring by checking if any of the image offsets are outliers when compared to all other offsets.
comment in response to
post
One roadblock that I've ran into is that my relatives are visiting so I can't use my 3D printers since that's the room they are staying in.
comment in response to
post
I've been working on my Lego sorting machine, and Forge's GUI and image stitching algorithm. For the Lego sorting machine, I designed and 3D printed custom sprockets and mounting plates to allow for mounting the sprockets on metal rods and to prepare for the next step of mounting the motor to it.
comment in response to
post
Hello, I'm Procerand and I'm making a game from scratch and work on electronics projects. My main projects that I'm currently working on is a multiplayer 2D sandbox game, a lego sorting machine, and a low cost gigapixel sample imager for lab use.
comment in response to
post
This is a computational problem, although comparing them is done visually since there is no ground truth data to compare them against.
comment in response to
post
I think this is the first GBC module I've seen that uses Duplo. Then again, I haven't been keeping up with the new modules people have been making.
comment in response to
post
In addition, I'll also be reimaging the samples I currently have to figure out what overlap will end up being the best in terms of speed and reliability.
comment in response to
post
Yeah, the lab gave me access to a bunch of the images used to create the stitched versions of some of their old samples which have ~30% overlap (compared to ~75% for Forge) so I'll be trying it on those.
comment in response to
post
As for my project for this upcoming week, I'll be designing and 3D printing sample slides for easy alignment, working on multi sample imaging. Get the motor mounts designed, and implement proper game state management in my game now that the multiplayer system is working.
comment in response to
post
I also confirmed that my image stitching algorithm for tree core samples produces the highest quality results out of all the software the lab has tried before, although it still needs to be tested on images with lower overlap.
comment in response to
post
I learned how to do parametric parts in Fusion 360 (super useful btw) which I then used to design custom sprockets for my Lego sorter. Unfortunately, I haven't gotten around to the motor mount yet. I also have been working on improving the GUI to Forge.
comment in response to
post
Here's the report: www.europol.europa.eu/cms/sites/de...
I can't find the quote or the statistic anywhere. I'm pretty sure it was invented whole-cloth by someone using AI to write blog posts. About how you can't trust what you read on the Internet. Which is now part of Google's AI summary.
comment in response to
post
Maybe they're talking about Chat GPT's advanced reasoning where it has the intermediate steps that aren't shown in the final answer? But that's specific to Chat GPT and not LLMs in general.
comment in response to
post
If I had to guess, I'd actually assume that code generation is actually just as intensive as generating normal text. The only difference is being the amount of tokens used as I'm fairly confident that it uses more tokens than normal text. So you'd need to break it down by cost per token to compare.
comment in response to
post
Another thing I just noticed is the D in day has a weird shadowy blob above it, and the straight lines on the side also have this fuzziness to them that all the other lines don't.
comment in response to
post
The apostrophe, the necklace, the eyes being closed, and the grain over the entire image and not just the background makes me think it's AI. Necklace kinda blends together at parts, and I've never seen an apostrophe that looks like that.
comment in response to
post
If you like that, you should checkout the Lumen PnP project made by @stephenhawes.com as he's working (and has released) an open source pick and place machine
youtu.be/JlkTcxh-9gA?...
comment in response to
post
I think I found it. It's called Pellet Stoneleigh and it's part of Norman Architecture. Unfortunately, I've only been able to find examples of it on church arches. I got the clue from an old book that someone has scanned and posted online.
comment in response to
post
I found something similar to the masonry patterns on the walls you've posted, but I feel like it doesn't quite fit entirely since it isn't only the spheres. There's two different techniques that I've been able to find "Egg and Dart" (top of image) and "Bead and Reel"(bottom of image)
comment in response to
post
Don't forget the, "Just follow this guide" with a link to a site that is no longer around and someone replies, "Thanks it worked perfectly!"
comment in response to
post
If you're talking about this, its literally just a cannon ball stuck in a wall.
comment in response to
post
Although now that I'm rereading it you said a spherical, not a round stone.
comment in response to
post
Like this right? This is called a Millstone Feature. If you look up In Wall Millstone Feature you should find more pictures.
comment in response to
post
I hope I can get back to the VR work in a reasonable amount of time as well.
As for the motor I'm going with the Greartisan 12v 50 rpm high torque motor, but I might switch to a different one at a later date if I find it to be too slow.