I discovered Image Creator today; which is a free online software obtained through Bing. I wanted to see what could be done by way of creating hard contrast images that will readily and easily convert to 3D images in Carveco Maker Plus. (I upgraded from Carveco Maker with a $301 discount for black Friday).
Image Creator is really easy to use and seems to be a pretty competent implementation of an AI image processor. The first sample image is a cobra snake about to strike, showing its fangs. My simple input text was as follows…
A grayscale image of a a cobra snake about to strike and showing its fangs. The image should be framed with the highest possible contrast, representing a relief model for CNC routing. The image must feature extremely balanced lighting to make details visible across the model, with very clear, sharp contours and textures. The design should have strong shadows and intense highlights to accentuate different levels and surfaces, ensuring no detail is obscured.
The user can choose from four suggested images. The frame was not to my taste and I removed it. My choice of image is this one:
By adjusting the input text one can create any subject… this image is a Siamese cat .It also shows great contrast and will be easily converted into a height map or an STL file.
With more understanding and better refinement of the input text by me, I could see that a much wider selection of models for 3D woodwork will become more readily available. This form of image creation is an exciting new tool in the armoury of all who would use CNC machining to create 3D models.
Your cat image is so serene and regal. However I know that cat is plotting to kill its owner and eat them. Cats are strange animals. We have lived with cats for thousands of years but they are still wild animals at heart.
Your images look very nice. Update us with something you carved with your images.
Thank you. It is heartening to learn the images do not look like they were created by a beginner. Yes, I will do so when I get to grips with 3D modeling in my upgraded Carveco Maker plus. I will create the cat first… I like cats and have never been killed by one yet. I had a friend for many years (now sadly passed) who kept 18 rare endangered leopards in his garden of a few acres.
Oftentimes I took my kids to see these gorgeous animals in their semi-wild environment. I recall that there was not much more than chicken wire between the leopards and visitors. The very rarest animal was a clouded leopard with markings looking like they belonged on a giraffe.
One strange by-product of my friend keeping these animals at his home was this. Trainee veterinary surgeons at London Zoo used to come and look after the leopards as an integral part of their training. Their services were provided free.
Just a couple of screenshot examples of the depth map and the .stl files. On this showing, I think there may be more detail in the depth map and it should provide me with a bit more room to adjust the final carve. More anon…
I’ve been using Bing Image Creator for about a year, mostly for laser work. I’ve also used Google Gemini, which is another free tool.
As you mentioned, it’s important to phrase your prompts correctly to get the desired image.
Here’s an image created with Gemini after copying and pasting your input text of the cobra. Exploring what can be made using these images with the router is on my radar for next year.
Thanks for the heads up and the illustration of what my sample input test returned on Gemini. I really like the orderly detailed and vignetted Gemini image. It could form the basis of a fantastic laser image with an adjusted central contrast range. I guess I will have trot on over to Gemini to see what is on offer.
I have used it a couple of times and when it presents the panel of 4 choices I liked one but wanted some changes. I would tell it to use that but add x or Y ow whatever and the resultant images were completely different.
Yes, software which offers you much more control tends not to be available for free. You can regard the issue of changing the file every time your request is received (while adhering to the input command text) as the failure to implement the structure required to accept a partially changed input, while leaving the rest of the file untouched.
The Image Creator AI appears to choose its own operating parameters every time, while broadly keeping in line with the terms of your input text command. Ideally, you would want something similar in structure to the pipe command in Unix. This is where the output file of one operation becomes the input file for the next operation. This type of program structure could permit you to make changes to the output file that is to be used as the input for the next operation.
I had a friend for many years (now sadly passed) who kept 18 rare endangered leopards in his garden of a few acres.
