Introduction to Image Formation



1. Overview of Image Formation

Image formation refers to the conversion of visual information from the physical world into a digital image. It is the first and foundational step for any computer vision or digital image processing task.

2. Stages of Image Formation

a) Physical Scene

  • The real-world environment or objects that are to be captured.



b) Illumination

  • Light plays a crucial role in how scenes appear.

  • Different sources (natural vs. artificial) affect color, brightness, and shadow.



c) Camera Optics (Lens System)

  • Lens: Focuses incoming light onto the sensor surface.

  • Aperture: Regulates the amount of light by changing its size.

  • Shutter: Controls the exposure duration—how long light hits the sensor.



d) Image Sensor

  • The sensor (CCD - Charge-Coupled Device or CMOS - Complementary metal-oxide-Semiconductor) turns photons (light) into electrical signals.

  • Each pixel on the sensor collects light and converts it into a voltage proportional to intensity.



e) Sampling and Quantization

  • The continuous image formed on the sensor is sampled into discrete points (pixels).

  • Each pixel's analog value is quantized (converted) into a digital number, e.g., 0–255 in 8-bit depth.



3. Key Steps: Flowchart Representation

StageWhat Happens
Physical SceneObjects exist in presence of light
IlluminationScene is lit by natural or artificial sources
Camera & LensLight is focused and managed onto the sensor
SensorLight intensity is converted into analog voltage
DigitizationAnalog values are sampled & quantized to digital
ImageDigital values form a pixel grid (the image)

4. Types of Digital Images

  • Grayscale – Each pixel represents brightness only (intensity).

  • Color (RGB) – Three channels; each pixel stores values for Red, Green, Blue.

  • Other formats – Multispectral, infrared, thermal, etc., depending on the sensor.



5. Mathematical Models

  • Pinhole Camera Model: Projects 3D world into a 2D image using simple geometry:

    x=fXZ,y=fYZ

    where ff is focal length; X,Y,ZX, Y, Z are real-world coordinates.



  • Lens Camera Model: Incorporates focus and distortion characteristics for realistic image models.



6. Influence Factors

  • Noise: Caused by sensor imperfections or low light.

  • Blur: Owing to camera movement or defocus.

  • Lens Distortion: Straight lines appear curved due to optical properties.




7. Applications & Importance

  • Understanding these steps is essential for any task in computer vision:

    • Pre-processing: Denoising, deblurring, color balancing depend on image formation knowledge.

    • Scene Understanding: Knowing formation helps with meaningful interpretation.

    • System Design: Camera choice, sensor type, and pre/post-processing all rely on these fundamentals.

8. Summary

Understanding image formation is crucial as all later processing depends on the quality and nature of the acquired image. Mastery of this topic ensures robust performance in computer vision, image enhancement, and analysis.