Discerning approximate distances of facial landmarks through improved computerized holography

Abstract:

Division of Facial Regions

Dividing up an image into groups is a common task in computer graphics. https://medium.com/mlcrunch/face-detection-using-dlib-hog-198414837945 this article talks about dividing the facial portrait into regions based on squares. This is why if you look at a Washington state ID, there are squares drawn over the ID picture.

Depth of Field

Depth of field (DOF) refers to the area of focus between the nearest and furthest objects in focus in a photograph. In an image with shallow DOF, only a slither is in focus. You have your lens at the largest aperture and you are focused on something ten feet away, everything that is ten feet away will be in focus. Anything that is more or less than ten feet away will not be in focus. You can think of focus as a large plane, in this case ten feet away from the camera. Or like a large sheet of glass parallel with the back of the camera. This is called the plane of focus. Anything intersecting that plane will be in focus. A research topic on "depth of field for a group of objects" could explore how the arrangement and distance between multiple objects within a scene affects the depth of field in a photograph, investigating factors like object spacing, focal length, aperture, and camera-to-subject distance, and how these elements can be manipulated to achieve desired focus effects when capturing a group of subjects in a single image. 

https://www.dofmaster.com/dofjs.html

Lessons from 3D sidewalk chalk

Here is an article about 3D sidewalk chalk: https://digital-photography-school.com/understanding-depth-field-beginners/  Depth perception arises from a variety of depth cues. These are typically classified into binocular cues and monocular cues. Binocular cues are based on the receipt of sensory information in three dimensions from both eyes and monocular cues can be observed with just one eye.^[2][3] Binocular cues include retinal disparity, which exploits parallax and vergence. Stereopsis is made possible with binocular vision. Monocular cues include relative size (distant objects subtend smaller visual angles than near objects), texture gradient, occlusion, linear perspective, contrast differences, and motion parallax.^[4]

Potential Research Directions

• How does the distance between objects within a group impact the overall depth of field? (e.g., comparing a tightly packed group vs. a widely spaced group) 
• What is the optimal camera position and focal length for maximizing the depth of field when photographing a group of objects at different distances? 
• How does aperture setting influence the depth of field when capturing a group of objects at varying distances from the camera? 
• Can depth of field be used creatively to emphasize specific objects within a group, while subtly blurring others? 
• How does the size and shape of the objects within a group affect the perceived depth of field in a photograph? 

Possible research methods:

• Controlled photography experiments:

  Setting up various group arrangements with different object distances and camera settings to analyze the resulting depth of field. 
• Image analysis software:

  Utilizing software to measure the sharpness of different areas within a photograph to quantify the depth of field. 
• User studies:

  Asking participants to evaluate the visual impact of different depth of field effects on group photographs 

Applications of this research could include:

• Portrait photography: Optimizing group portrait composition to ensure all subjects are in focus. 
• Product photography: Arranging multiple products in a way that highlights the desired focal point while maintaining acceptable sharpness across the group. 
• Landscape photography: Capturing scenes with foreground and background elements in focus, depending on the desired artistic effect

Using Holographic Tools as related to Depth of Field

A research topic related to hologram depth of field could focus on developing techniques to significantly increase the depth of field in holographic imaging systems, allowing for sharper 3D reconstructions across a wider range of object distances, which is currently a major limitation in holographic technology; this could involve exploring methods like computational holography include incorporating deep learning algorithms to optimize the reconstruction process and extend the depth of focus. 

Key areas of research within this topic could include:

Computational holography with optimized algorithms:

Designing algorithms that can generate holograms with a wider depth of field by manipulating phase information and utilizing advanced computational techniques.

Light-field holography:

Investigating the integration of light-field imaging principles with holography to capture and reconstruct 3D scenes with extended depth of field.

Applying deep neural networks to analyze and reconstruct holographic data, potentially enabling real-time depth of field adjustments. 

Potential applications of this research:

Advanced augmented reality (AR) displays:

Creating more realistic and immersive AR experiences by overcoming the depth of field limitations of current holographic displays. 

Medical imaging:

Improving the quality of 3D holographic visualizations of medical scans by extending the depth of field. 

Industrial inspection:

Utilizing holographic imaging with greater depth of field for detailed 3D analysis of complex objects.

 

Works referenced

https://docs.real.com/guides/safrdocs/Camera_Best_Practices.html

https://skullsinthestars.com/2008/07/14/in-depth-depth-of-focus-and-depth-of-field/

https://fxhome.com/blog/what-is-depth-of-field

https://www.edmundoptics.com/knowledge-center/application-notes/imaging/depth-of-field-and-depth-of-focus/

https://en.wikipedia.org/wiki/Holography

https://opg.optica.org/oe/fulltext.cfm?uri=oe-31-3-4199&id=525412#:~:text=Abstract,%2Dthe%2Dart%20DFF%20methods