Imagine you've made a concept drawing on paper and you want to use it as a reference for a 3D model. Or maybe you're better at drawing with a pencil than you are at drawing with a tablet, and you want to use one of your sketches as game art. You can use a scanner to get your art into your computer, but where I live scanners cost AU$100 or more. Instead, I will outline a method for using any digital camera — even a smartphone — to scan a drawing at high detail, with consistent and reproducible scale and clarity.
The image used here was drawn on grey watercolor paper, scanned with this method, and then ten minutes was spent touching it up in GIMP (background removed using Fuzzy Select, colors adjusted).
The final result, suitable for placeholder art or direct use.
Why Not Just Photograph It?
We all have access to digital cameras now, even in our smartphones. We can just take a picture of the drawing and load it up in the computer, right? Well, no. It's hard to get consistent results from just pointing your camera and shooting:
- It's quite difficult to hold it steady and avoid blurring.
- The brightness of your photos will differ.
- The viewing angle will not be perpendicular, so your drawing will be skewed.
- The distance from the drawing to the lens will differ, so drawings won't be consistently scaled.
- The lighting in the room may be uneven and create dark spots, or may colorise the drawing (taking a photo in sunlight will be different from taking it by lamp light, for example).
- Capturing the drawing with a single photo removes the fine detail of the drawing. You may want to capture the texture of crayon on rough paper, or maybe your concept art is very intricate.
You need a way to control all of these variables.
Simply taking a photo of the drawing can produce unwanted shadowing and a skewed perspective, leading to more time spent touching it up digitally.
Solution: Make a Composite Image
A composite image (one built from many smaller images) is one way to get a high-detail reproduction from a digital camera. The method I outline here uses a free Windows tool called Microsoft Composite Image Editor to automatically stitch photos together with no extra work on your part. All you need to worry about is getting a set of consistent photos, and you do that by making a guide.
Step 1: Make a Guide
A guide is simply a platform that holds your camera in a consistent way. You can make it out of whatever you want, as long as it fulfills these requirements:
- Holds your camera perfectly steady.
- Keeps the camera level, and at the same distance from the drawing at all times.
- Allows external light in, diffuses the light evenly, and is white to avoid coloring the drawing.
I made my guide from a translucent plastic soup bowl, the kind you get from a takeaway Chinese restaurant. The bowl has a stiff rim around the bottom, so I simply cut the entire bottom out without losing any rigidity.
Step 2: Add Light
Set up your drawing and guide in a well-lit place - the more light the better. Beside a sunny window is excellent. You can also use a white light at the side of the container to light the entire thing. My white bowl diffuses the light fairly evenly throughout the interior.
This light source is a simple USB-powered LED lamp from eBay.
Step 3: Take Photos
Place your camera on the guide and move the guide over the drawing, taking a series of photos that overlap by a comfortable margin. For best results, try not to rotate the camera.
Notice how much overlap I included to give the compositing program plenty of clues.
Step 4: Load the Images Into Microsoft ICE
Pull your images straight off your camera and drag-and-drop them into Microsoft Image Composite Editor. Download it here if you haven't already. If your pictures have been taken correctly, ICE will detect the camera motion as Planar, and stitch the images together into a single image. If ICE doesn't do this, check that your camera focus is right, and that the lighting is consistent between images.
Step 5: Export the Composite Image
Once you've exported your image, you can use your favourite image editor to modify it as necessary. For simple line art (outlines for pixel art, for example) you might desaturate the image down to three or four colors to isolate the darkest lines. You can raise saturation if your lighting washed out the colors, and you can use Fuzzy Select to turn areas of plain paper into transparency. Experiment!
The final result, with background cleaned up and colors intensified.
If you have a copy of Adobe Illustrator, check out this Vectortuts+ tutorial about creating line art from a scanned sketch.