Metadata-Version: 2.1
Name: pic2world
Version: 0.4.1
Summary: P2W is a toolbox that implements several utilities for getting real-world information from pictures.
Home-page: https://github.com/Eric-Canas/pic2world
Author: Eric Canas
Author-email: eric@ericcanas.com
License: MIT
Platform: UNKNOWN
Description-Content-Type: text/markdown
License-File: LICENSE

# Pic2World
<b>Pic2World</b> is a toolbox for inferring <b>Real World Information</b> from <b>Pictures</b> and <b>Camera Intrinsics</b>.
It also offers a set of <b>Geometry</b> and <b>Image Manipulation</b> tools, that help to deal with <b>Homographies</b>
and <b>Perspective</b> while maintaining the <b>Real World</b> coherence.

## Installation
You can install Pic2World [via pip](https://pypi.org/project/pic2world/), by running the following command:
```bash
pip install pic2world
```

## Usage
Pic2World includes 4 main modules:
1. `pic2world.geometry`: Geometry tools for dealing with polygon transformations.
2. `pic2world.camera_utils`: Tools based on [Gauss Thin-Lens Equation](https://en.wikipedia.org/wiki/Thin_lens) for measuring real world distances from pixel measures and camera intrinsics.
3. `pic2world.homographies`: Tools for dealing with Homographies. Designed to produce visualizations coherent with Real World inferred data.
4. `pic2world.interactive`: Tools for interactive visualization, based on [matplotlib](https://matplotlib.org/). Help to define polygons and ask questions to the user. Useful for debugging.

### Camera Utils
Camera Utils include two main classes, `pic2world.camera_utils.camera` and `pic2world.camera_utils.ruler`.


`Camera` is used to contain all Camera Intrinsics information:
```python
from pic2world.camera_utils.camera import Camera
# Build the Intrinsics of a Canon R6 
CANON_EOS_R6_CAMERA = Camera(
    pixel_size_mm=8.18/1000.0, # Pixel size of the camera sensor
    focal_length_mm=50, # Focal length of the mounted lens
    sensor_shape_px=(3648, 5472) # (height, width) of the sensor. Can be setted as None. Only used when images are taken from hard angles.
)
```

`Ruler` is used to calculate real world distances from pixel measures and camera intrinsics:

##### Calculating Distance between the lens and the object when the Real Length is known


```python
from pic2world.camera_utils.ruler import Ruler
# Create a Ruler object
ruler = Ruler(camera=CANON_EOS_R6_CAMERA)
distance_to_img = ruler.distance_to_object_cm(object_length_px=2320, # Length of the object in pixels
                                              real_object_length_cm=3*30.0, # Real Length of 3 DIN-A4 papers.
                                              angle_degrees=60) # Angle with which the image was taken (0 would mean zenith).
```

```python
# Print the distance in cm.
>>> Distance to object -> Calculated: 118.5608 cm [Real: 118.0 cm]
```

##### Calculating Real Length of an object when the distance between it and the camera lens is known

```python
# Assume the same camera as above
object_height = ruler.object_length_in_cm(object_length_px=2320, # Length of a vertical of the object in pixels,
                                                  distance_to_object_cm=118.0, # Distance between the object and the camera lens in cm
                                                  angle_degrees=60) # Angle with which the image was taken (0 would mean zenith).
```

```python
# Print the real length in cm.
>>> Object height -> Calculated: 89.5742 cm [Real 90.0 cm]
```

### Homographies
Homographies module includes functions for changing the perspective of an image, while maintaining the real world coherence.

##### From original image to zenith view. Setting input interactively

```python
from pic2world.homographies.homography_utils import correct_perspective
correct_polygon_perspective(img,
                            origin_polygon=None, # We are not providing an input polygon because we want the user to define it.
                            interactive=True, # Ask the user to define the polygon.
                            angle_degrees=60.0, # Angle with which the image was originaly taken.
                            output_shape=(600, 300), # Output shape we want
                            pad= 0.05) # Padding between the limits of the rectangle and the border of the output image.
```

## Note

This library is a work in progress. It is not yet complete, and it is not meant to be used in production yet.

