2D to Pseudo-3D (Customizable)

thingiverse

This process revolves around creating a pseudo-3D model, or "lithophane," from an image based on its gray values. Using Matlab and ImageJ, you can achieve this without requiring any light source. To begin, import your desired RGB image into Matlab by simply dragging and dropping it (e.g., "image2D.jpg"). Next, cluster the image using the "clusterImage" function with two parameters: the 2D matrix of the image and the number of desired clusters, applying k-means algorithm to create a clustered image based on gray values (e.g., "clusteredImage = clusterImage(image2D,16);"). To construct the pseudo-3D model, utilize the "calcPseudo3dModel" function with two parameters: the clustered image from step 2 and the desired number of depth layers (e.g., "pseudo3dModel = calcPseudo3dModel(clusteredImage ,32)");. Then, save the created pseudo-3D model as a sequence of images using the 'savePseudo3dModel' function with three parameters: the pseudo-3D model, a file name (e.g., "semih"), and a flag that determines inversion (e.g., "0");. Next, import the image sequence into ImageJ and open the 3D Viewer, selecting 'Display as: Surface', 'Color: White,' and 'Resampling factor: 1'. Export the model as an STL file from File->Export surface as-> STL(binary). Lastly, open the STL file using a software like Repetier and adjust the size of your lithophane model as desired.