I thought I would tap into the extensive knowledge of Form users:)
I have a (potential) client that has a series of images (1000 to be exact). Each image is 1 layer of an object. So each image stacks on top of each other to create a 3D object. Think of 1 sheet of paper and then 1000 sheets of paper together.
Currently the images are in png format, does anyone have any idea how these could be used to create a 3d object that could then be printed?
1000 PNG images as layers? That sounds like the output of program that controls a projector-based printer. I don’t know of any way to turn that back into STL. And the fact he has those but not the STL sounds to me like he doesn’t have the rights to the model and just got his hands on this.
Nope, he is a microscopists. So for further information to put this in context: he has datasets that exist as a sequence of images recorded using a technique known as ‘serial block-face scanning electron microscopy’. This involves cutting away layers of an object using a diamond knife, and recording an image of each slice in an electron microscope. These datasets can be quite large, a typical example being 3000 images of dimension 3000 x 3000 pixels.
Perhaps people in his field know of software? I imagine there needs to be different models created from different sets if lines, otherwise the printer will just print a solid block . If there’s a need for this and you can’t find anything existing, the last resort would be to hire a programmer.
If you really needed to do it manually, you could use Illustrator to trace the images and then export the paths to a 3D program and just convert each layer to a mesh, it would take a bit of monotonous work but it’s possible.
I have imaged both with MRI and CT machines and convert them to STL files. The micro-CT, which I have used, produces data like yours as a folder of all of the ‘slices’ in either jpg, png or bitmap formats. I then used a medical imaging program, like MIPAV (free) to convert this to a more familiar 3D Cartesian image matrix. This conversion is relatively quick and painless. From there I can read it into another program, ITK-Snap to segment out a volume, or the entire image. This volume ‘segmented’ image will be converted to the STL format, which is a surface rendering, and can be time and computation intensive, depending upon the goal of the data set extraction. There are other programs like this to perform the same functions. I clean things up then in NetFABB, MeshLab, and Meshmixer, as each is better at some functions than others. Then I’m ready to print.