Greyscale maps are 2D images where each pixel represents a single value corresponding to a shade of grey that has a certain intensity or brightness. In an average greyscale image, each pixel is an 8-bit value ranging from 0 (black) to 255 (white), with the intermediate values represent various shades of grey. Factum typically uses a 16-bit greyscale file, representing 32,767 different shades of grey. This enables a higher legel of detail to be achieved and smoother gradients in comparison to the 8-bit greyscale image. These 2D maps are commonly used for 3D visualisation, image processing, and geological surveys and increasingly AR/VR.
The Shape from Shading (SfS) technique developed by Horn (1989) forms the basis for estimating the 3D shape of a surface from a 2D greyscale image. This uses the variations in brightness intensity, or shading to unveil surface normal information and infer the three-dimensional shape of objects. Imagine a photograph of a human face. Even though the reflective properties of the skin are consistent, brightness variations are evident. Why? These variations arise because different parts of the surface are oriented in such a way that they reflect the incoming light to the viewer differently.
Greyscale maps are the result of a long history of cartography and artistic representation techniques. In the realm of cartography, they share similarities with Digital Terrain Models (DTM) as both use shades of grey to represent terrain data. However, DTMs are a specific type of digital representation that focuses on elevation information (slope calculations, hydrological analyses), while greyscale maps can represent various types of data in a simplified, monochromatic form.
A Digital Terrain Model (DTM) can be represented in several ways:
Factum uses recording technologies (Lucida and Selene) that store high-resolution data as greyscale depth maps, where the various shades of grey correspond to a particular height on the scanned surface, thereby allowing the object's three-dimensionality to be represented in a two-dimensional image.
Here, illumination transcends being a mere visualisation tool, instead, it acts as a beacon that uncovers and enriches our understanding of these objects. Through the variations of light and shadow three-dimensional data is recorded with the utmost precision.
While greyscale maps provide a great foundational analysis of an object’s topology, incorporating other data visualisation methods and technologies can further enhance our understanding and appreciation of material culture under observation. Greyscale can be complemented by capture methods like Shape from Motion (better known as photogrammetry). These varied techniques mean that the capture protocol leverages the collective capabilities and strengths of several methods and ensures comprehensive and detailed documentation, capturing not just the visible, but also the minutiae that might elude a single-method capture.
On the other hand, depth maps are directly generated or measured from a 3D scene or object, and each pixel value represents the depth information. While SfS techniques can provide depth estimation indirectly from greyscale images, depth maps themselves are not typically used as inputs for SfS algorithms.
A image of a greyscale depth map.
A greyscale image normally has values for each pixel ranging from 0 (representing absolute black) to 255 (symbolizing pure white). This range of 256 shades of grey corresponds to a bit depth of 8, meaning that 8 bits are used to store the value of each pixels. An 8-bit depth image contains enough information for common applications. However, in the context of surfaces with complex and subtle height variations, an 8-bit height map is not suitable as there are not enough depth levels and so vital details are lost.
To solve this problem, at Factum we generate 16- and 32-bit depth images (TIF files) using our bespoke technologies; The Lucida and The Selene. 16- and 32-bit depth images capture an increased number of altitude or depth values that can be assigned to each pixel. The Lucida and Selene generate output files with 32 bits of resolution (storing values with a resolution of up to 0.0000001 meters). This extreme precision and accuracy of surface height and depth are essential for faithfully representing the subtleties and complexities of the object recorded.
In Factum's workflow, the transition from 32 bits to 16 bits occurs when we materialise a digital record into a facsimile. This is because although 32 bits provide maximum fidelity, 16 bits become more than suitable when we rematerialise these objects as facsimiles.
The main benefit of using greyscale maps to hold surface information is owing to ability to store the ultra-high-resolution data efficiently (recording height differences below 100 microns). We store the raster image data as TIF files, which are also compatible with various Geographic Information System (GIS) tools but also with image editing programs such as Adobe Photoshop. Thereby enhancing the interoperability with other software tools and systems.
The intersection of cartography and cultural heritage preservation through greyscale depth maps illustrates how technology can be adaptive and interdisciplinary. By borrowing from the science of mapping our world, we've enriched the science of preserving it.
