A simple method is to shoot from exactly the same position with different lighting conditions so that the shadows appear at different positions. Then you align the images (even though you are shooting from the same fixed position, a shift of a few pixels can easily occur), and transform to linear colorspace. A linear combination of these pictures then corresponds to how the picture would have looked if you had taken it under the corresponding linear combination of the different lighting conditions. You must make sure the white balance is fixed, you should not use auto-white balance.
This allows you to compile a picture of the object illuminated by many lights, shining from different direction so that it is almost uniformly illuminated. When you are done, you transform the result back to sRGB colorspace and then you can correct the white balance.