Gamma correction

Gamma correction, also referred to as gamma encoding, is an image processing technique that adjusts the brightness and contrast of an image to achieve a more natural and visually pleasing appearance.

  • Gamma correction of digital images prevents excessive storage of information about highlights that are invisible to humans and ensures sufficient information is retained for shadows that require more differentiation to be clearly observed.
  • Gamma correction adjusts the relationship between the numerical value of a pixel stored in an image file (e.g., JPG or TIFF) and its corresponding brightness when displayed on-screen.
  • Gamma correction is typically performed to compensate for the non-linear relationship between the input signal and the displayed brightness on a monitor or screen.
  • In the case of a black-and-white image, a gamma function impacts highlights (brightest values), mid-tones (greyscale), and shadows (dark areas) in distinct ways.
  • Gamma correction is not limited to black and white images but applies to colour images, where it affects colour balance and contrast.
  • The appearance of an image on a digital display is determined by the voltage associated with each pixel:
    • For instance, a computer utilized to display black-and-white images translates the numerical values of each pixel in an image file into a corresponding voltage, which is then transmitted to a monitor. The brightness of a pixel increases with higher voltages.
    • The ideal relationship between stored values and appearance is non-linear, meaning that a voltage change does not directly result in a satisfactory brightness change from an observer’s perspective.
    • For many TVs and computer displays, doubling the voltage of a specific pixel will not make it appear twice as bright. Therefore, gamma correction selectively adjusts voltages to enhance the overall appearance.
  • Gamma correction can help achieve accurate representation of images across various display devices and ensure consistent visual experiences.
  • Different applications and devices may have different default gamma settings, and users can often customize these settings based on their preferences.

Gamut

The term gamut or colour gamut can be used to describe:

  • The range of colours that a specific device or system can display or reproduce.
  • The range of colours that the human eye can see in specific conditions.
  • A range of colours smaller than all the colours that the human eye can see.
  • All the colours in an image. Digitising a photo, changing an image’s colour space, or printing it might change its gamut.
  • The range of perceived colours (visible to a human observer) is always greater than the range that can be reproduced by any digital device such as a screen, monitor or projector.
  • Digital cameras, scanners, monitors, and printers all have limits to the range of colours they can capture, save, and reproduce.
  • The main use of digital colour spaces and colour profiles is to set the gamut of colours that can be used to accurately reproduce or optimise the appearance of an image.
  • It is currently impossible to make a digital device that can reproduce the same range of colours that the human eye can see.

Ganglion cells

A retinal ganglion cell is a type of neuron located in the retina of the human eye. It receives visual information from photoreceptors via two intermediate types of neurons (interneurons): bipolar cells and retina amacrine cells.  Retinal ganglion cells transmit image-forming and non-image-forming visual information to several regions in the thalamus, hypothalamus and midbrain.

  • Retinal ganglion cells are located near the boundary between the retina and the central chamber containing vitreous humour.  They collect and process all the visual information gathered directly or indirectly from the forty-something types of rod, cone, bipolar, horizontal and amacrine cells and, once finished, transmit it via their axons towards higher visual centres within the brain.
  • The axons of ganglion cells form into the fibres of the optic nerve that synapse at the other end on to the lateral geniculate nucleus. Axons are like long tails and typically conduct electrical impulses, often called action potentials, away from a neuron. They take the form of long slender fibre-like projections of the cell body.
  • A single ganglion cell communicates with as few as five photoreceptors in the fovea at the centre of the macula. This produces images containing the maximum possible resolution of detail. At the extreme periphery of the retina, a single ganglion cell receives information from many thousands of photoreceptors.
  • Around twenty distinguishable functional types of ganglion cells resolve the information received from 120 million rods and cones into a million parallel streams of information about the world surveyed by a human observer throughout every day of their lives. Their functions complete the construction of the foundations of visual experience by the retina, ordering the eye’s response to light into the fundamental building blocks of vision.  Ganglion cells enable retinal encodings to ultimately converge into a unified representation of the visual world.
  • As described above cone cells are attuned to different bands of wavelengths, with peak biases at 560 nm, 530 nm, and 420 nm and are concerned with trivariance – three discernible differences in the overall composition of visible light entering the eye.
  • Ganglion cells also play a critical role in trichromacy but the way they function might be thought of as being determined by limitations on bandwidth within the optic nerve.
  • Ganglion cells not only deal with colour information streaming in from rod and cone cells in real time but also with the deductions, inferences, anticipatory functions and modifications suggested by bipolar, amacrine and horizontal cells. Their challenge, therefore, is to enable all this data to converge and to assemble it into high fidelity, redundancy-free, compressed and coded form that can continue to be handled within the data-carrying capacity of the optic nerve.
  • It is not hard to imagine the kind of challenges that have to be dealt with:
    • The information must feed into and support the distinct attributes of visual perception and be available to be resolved within the composition of our immediate present visual impressions whenever needed.
    • The information must correspond with the outstanding discriminatory capacities that enable the visual system to operate a palette that can include millions of perceivable variations in colour.
    • Information about the outside world must be able to be automatically cross-referenced, highly detailed, specifically relevant, spatial and temporally sequenced and available on demand.
    • The information must be subjectively orientated in a way that it is locked at an impeccable level of accurate detail to even our most insane intentions as we leap from rock to rock across a swollen river or dive from an aircraft wearing only a wingsuit and negotiate the topography of a mountainous landscape speeding past at 260km per hour.
  • It is now known that efficient transmission of colour information is achieved by a transformation of the initial three colour mechanisms of rods and cones into one achromatic and two opponent chromatic channels. Opponent type processing clearly represents the optimal necessary step to dynamically readjust the eye’s earlier trivariate responses to meet the criteria of efficient colour information complete with all the necessary contextualising detail ready for transmission. We can assume it is in response to these demands that every stimulus to the eye can be accurately and objectively defined in both space and time in ways relevant to everyday circumstances.

Ganglion cells

Ganglion cells

Retinal ganglion cells are located near the boundary between the retina and the central chamber containing vitreous humour. They collect and process all the visual information gathered directly or indirectly from the forty-something types of rod, cone, bipolar, horizontal and amacrine cells and, once finished, transmit it via their axons towards higher visual centres within the brain.

The axons of ganglion cells form into the fibres of the optic nerve that synapse at the other end on the lateral geniculate nucleus. Axons take the form of long slender fibre-like projections of the cell body and typically conduct electrical impulses, often called action potentials, away from a neuron.

A single ganglion cell communicates with as few as five photoreceptors in the fovea at the centre of the macula. This produces images containing the maximum possible resolution of detail. At the extreme periphery of the retina, a single ganglion cell receives information from many thousands of photoreceptors.

Around twenty distinguishable functional types of ganglion cells resolve the information received from 120 million rods and cones into one million parallel streams of information about the world surveyed by a human observer in real-time throughout every day of their lives. They function to complete the construction of the foundations of visual experience by the retina, ordering the eyes response to light into the fundamental building blocks of vision. Ganglion cells do the groundwork that enables retinal encodings to ultimately converge into a unified representation of the visual world.

Ganglion cells not only deal with colour information streaming in from rod and cone cells but also with the deductions, inferences, anticipatory functions and modifications suggested by bipolar, amacrine and horizontal cells. Their challenge, therefore, is to enable all this data to converge and to assemble it into high fidelity, redundancy-free, compressed and coded form that can continue to be handled within the available bandwidth and so the data-carrying capacity of the optic nerve.

It is not hard to imagine the kind of challenges they must deal with:

  • Information must feed into and support the distinct attributes of visual perception and be available to be resolved within the composition of our immediately present visual impressions whenever needed.
  • Information must correspond with the outstanding discriminatory capacities that enable the visual system to operate a palette that can include millions of perceivable variations in colour.
  • Information about the outside world must be able to be automatically cross-referenced, highly detailed, specifically relevant, spatial and temporally sequenced and available on demand.
  • Information must be subjectively orientated in a way that it is locked at an impeccable level of accurate detail to even our most insane intentions as we leap from rock to rock across a swollen river or dive from an aircraft wearing only a wingsuit and negotiate the topography of a mountainous landscape speeding past at 260km per hour.

It is now known that efficient transmission of colour information is achieved by a transformation of the initial three trivariant colour mechanisms of rods and cones into one achromatic and two chromatic channels. But another processing stage has now been recognised that dynamically readjusts the eye’s trivariant responses to meet criteria of efficient colour information management and to provide us with all the necessary contextualising details as we survey the world around us. Discussion of opponent-processing is dealt with in the next article!

Geometric raindrop

In idealised terms, a raindrop is often represented as a geometrically perfect sphere. This simplification aids in comprehending the physics of rainbows, even though real-life raindrops seldom maintain such perfect spherical forms.

  • The understanding derived from studying the idealised geometry of raindrops can be applied to every rainbow despite the fact that:
    • The shape of a raindrop is highly variable and depends on factors including size, speed of descent, and turbulence.
    • Each rainbow observed in our daily life and the arrangement of droplets within it is unique due both to chance and to a wide range of environmental factors.
    • By way of summary, the form of a rainbow and the arrangement of raindrops within it depends on a variety of unique and changing conditions. These include the size, shape, and arrangement of the raindrops that make up the rainbow, as well as the position of the sun, the observer’s location, the clarity and composition of the atmosphere, and the presence of any other light sources or reflective surfaces. So, each rainbow that we observe is unique, shaped by both random variations and a wide array of environmental factors.
  • The idea of light rays is also a way to simplify the way we think about the behaviour of light as it approaches, passes through and exits raindrops towards an observer.
    • In reality, the notion of light rays does not describe an inherent physical property of light; rather, it’s a simplification for illustrative purposes.
    • More precise descriptions of light refer to it as composed of particles called photons, or as exhibiting wave-like properties.

Geometric raindrops

An idealised raindrop forms a geometrically perfect sphere. Although such a form is one in a million in real-life,  simplified geometrical raindrops help to make sense of rainbows and reveal general rules governing why they appear.

The insights that can be gained from exploring the geometry of raindrops apply to every rainbow, whilst the rainbows we come across in everyday life demonstrate that each individual case is unique.

Don’t forget that the idea of light rays is also a way to simplify the behaviour of light:

  • The idea that light is made up of rays is so commonplace when describing and explaining rainbows that it is easily taken for granted.
  • The idea of light rays is useful when trying to model how light and raindrops produce the rainbow effects seen by an observer.
  • Light rays don’t exist in the sense that the term accurately describes a physical property of light. More accurate descriptions use terms like photons or waves.
Basics of raindrop geometry
  • A line drawing of a spherical raindrop is the starting point for exploring how raindrops produce rainbows.
  • The easiest way to represent a raindrop is as a cross-section that cuts it in half through the middle.
  • A dot or small circle can be used to mark the centre whilst the larger circle marks the circumference.
  • Marking the centre makes it easy to add lines that show the radius and diameter.
  • Marking the centre also makes it easy to add lines that are normal to the circumference.
  • A normal (or the normal) refers to a line drawn perpendicular to and intersecting another line, plane or surface.
  • A normal is used in a diagram to connect the centre with a point where a ray strikes the circumference.
  • The diameter of a circle is a line that passes through its centre and is drawn from the circumference on one side to the other.
  • The radius of a circle is a line from the centre to any point on the circumference.
  • The horizontal axis of a raindrop is a line drawn through its centre and parallel to incident light. The vertical axis intersects the horizontal at 900 and also passes through the centre point.
  • The angle at which incident light strikes the surface of a raindrop can be calculated by drawing a line that shows where an incident ray strikes a droplet and then drawing the normal. The angle of incidence is measured between them.
  • The path of light as it strikes the surface and changes direction as it is refracted at the boundary between air and water can be calculated using the Law of Refraction (Snell’s law).
  • When light is refracted as it enters a droplet it bends towards the normal.
  • The law of reflection can be used to calculate the change of direction each time light reflects off the inside surface of the raindrop.
  • When light exits a raindrop the angle of refraction is the same as when it entered but this time bends away from the normal.

Geometrical optics

About geometrical optics
  • Geometrical optics, also known as ray optics, is one of the two main branches of optics, the other being physical optics.
  • Geometrical optics is based on the assumption that light travels as a straight line and is useful in explaining various optical phenomena, including reflection and refraction, in simple terms.
  • Geometrical optics is a useful tool in analyzing the behaviour of optical systems, including the image-forming process and the appearance of aberrations in systems containing lenses and prisms.
  • The underlying assumptions of geometrical optics include that light rays:
    • Propagate in straight-line paths when they travel in a uniform medium.
    • Bend, and in particular, refract, at the interface between two dissimilar media.
    • Follow curved paths due to the varying refractive index of the medium.
    • May be absorbed as photons and transferred to the atoms or molecules of the absorbing material, causing the absorbing material to heat up or emit radiation of its own.

Gravitational force

Gravity, the gravitational force, is one of the four fundamental forces in nature. The other forces are the electromagnetic force, the weak nuclear force and the strong nuclear force.

  • Gravity is the phenomenon that attracts objects with mass or energy towards one another.
  • It affects celestial bodies such as planets, stars, galaxies, and even light.
  • The influence of gravity on smaller objects like human beings in the presence of larger ones, such as planets, is evident.
  • Gravity, such as the Moon’s gravity, leads to ocean tides on Earth.
  • Gravity accounts for the weight of physical objects. Its range is infinite, although its effects weaken as objects move farther apart
  • Gravitational force is a universal force, meaning that it acts between all objects with mass, regardless of their composition or charge.
  • Gravitational force is a long-range force, meaning that it can act between objects that are very far apart.
  • Einstein’s theory of special relativity showed that mass and energy are equivalent, and can be converted into each other. This is expressed in the famous equation E = mc2, where E is energy, m is mass, and c is the speed of light.
  • This means that any object with energy also has mass, and therefore can be attracted by gravity. For example, light has energy, and therefore has mass. This is why light can be bent by very large objects such as galaxies.

Summary

Greyscale colour model

A greyscale colour model represents an image or graphic using shades of grey, with no colour information.

  • The greyscale colour model is used for:
    • Creating black-and-white images by cameras, scanners, and other input devices.
    • Converting colour images to black-and-white.
  • Three algorithms, namely the lightness method, weighted average method, and luminosity method, are used for greyscale conversion.
  • A greyscale colour model is not a linear scale from black to white but a way of converting colour brightness to show tonal relationships.
  • Converting digital images to greyscale involves assigning every pixel the correct level of brightness.
  • When fully saturated spectral colours are converted to greyscale, their brightness is usually between 11% and 89%. So:
    • Red = 70%
    • Orange = 40.38%
    • Yellow = 11%
    • Green = 41%
    • Blue = 89%
    • Violet = 74.06%
  • Any RGB decimal colour value can be converted to greyscale, so the decimal colour value corresponding with cyan is 178, 178, 178.
  • Any HSB colour value can be converted to greyscale, with the HSB value for pure yellow being Hue = 0, Saturation = 0, Brightness = 11.00%.
  • Greyscale images have many shades of grey in between black and white.
  • Greyscale images are distinct from one-bit bi-tonal black-and-white images which, in the context of computer imaging, are images with only two colours: black and white.
  • Greyscale images can be the result of measuring the intensity of light at each pixel against a selected wavelength or a weighted combination of wavelengths. In this case, the selected wavelengths can be from anywhere within the electromagnetic spectrum (e.g. infrared, visible light, ultraviolet.).
  • A colourimetric (or more specifically photometric) greyscale image is an image that has a defined greyscale colourspace, which maps the stored numeric sample values to the achromatic channel of a standard colour-space, which itself, is based on measured properties of human vision.