# Photogrammetry Theory

Photogrammetry is a process that converts a series of 2D photos into 3D spatial data using a camera and software. This section will introduce the thought process that we encourage one to adopt when approaching a photogrammetry project, as by understanding this theory will maximise your success.

## From 2D to 3D

The first part of the theory is to understand how 2D photos can generate 3D spatial data. The human eye and brain are able to process images and stitch them together into 3-dimensional information.

However, a 2D image does not contain any information on depth. We generally can *interpret* a photo in 3D because of context, but a computer (the piece of software) is unable to understand depth from an image. We perceive depth of field because our brain can interpret the parallax difference between the two different 'images' given to our eyes. Combined with our ability to move, we are able to observe a subject or environment from all its angles and understand its geometry and shape in 3D.

Photogrammetry relies on a similar process to understand 2D images as a 3D spatial data, it needs to create an understanding of depth. Similar **feature points** in the images are identified and extrapolated into 3D space to recreate the subject, this is called **triangulation.**

![Triangulation; feature points are extrapolated from the camera in 3D space to fine their depth](/files/-M2I7E-WOo48l3It2r1n)

## **Photo Alignment**

Generating 3D spatial data from a set of images is simple for a computer. 

The main challenge for the software is to align the input information. **Alignment** is the correct placement of images in 3D space for triangulation.

Imagine if you were given a set of images of a cube like below, and asked to take the exact same photos again, replicating the exact angles. Without any defining features to help you orient the photos as reference, it would be impossible to tell which angle each photo was taken from and at what distance:

![It is difficult to align these photos in 3D space. Simplifying for the software is the objective.](/files/-M2I7GGwqAQnLJQj4zWy)

## **Providing Features and Information**

The primary task of photogrammetry is to give the software sufficient feature points and consistent information for successful alignment of photos, which then can be triangulated into accurate 3D spatial data. 

These features and information can be bolstered in many ways and is covered in our Benchmark Guide. This Benchmark will cover the qualities to look out for in any subject and what techniques to apply to increase the amount of features captured.

{% content-ref url="/pages/-M2I8P2O1dltXE5tjGCJ" %}
[Photogrammetry Benchmark](/next-lab/3d-scanning/guides/photogrammetry/photogrammetry-benchmark.md)
{% endcontent-ref %}

To begin with photogrammetry, proceed to the Next page below: Photogrammetry Process.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://ms-kb.msd.unimelb.edu.au/next-lab/3d-scanning/guides/photogrammetry/photogrammetry-theory.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.