Temporal Filtering (2.2.1)

What is temporal filtering, and how does it enhance image contrast?

Temporal filtering involves averaging a sequence of images of the same scene to suppress noise and enhance contrast. It assumes that the scene is unchanging and the noise is mean zero and uncorrelated [a].

What are piecewise constant and piecewise continuous assumptions in temporal filtering?

Piecewise constant means the signal (e.g., pixel intensity) remains unchanged over intervals, like a stationary object in a video. Piecewise continuous allows smooth, gradual changes without abrupt jumps, such as a ball slowly moving across a scene.

What does the formula for noise in temporal filtering represent?

The formula [a] represents that the observed signal (Y_i) is the sum of the true signal (X_i) and additive noise (Z_i), where the noise is Gaussian with mean zero and variance σ².

How does temporal averaging reduce noise in image sequences?

By averaging N noisy images, temporal filtering reduces the variance of the noise from σ² to σ² / N [a].

What combination of techniques can recover details in low-light images?

Temporal filtering reduces noise, followed by histogram equalization to redistribute intensities. Together, they produce a clearer, high-contrast image.

How can temporal filtering assist in tasks like Optical Character Recognition (OCR)?

Temporal filtering reduces noise in image sequences, producing a cleaner image that improves the accuracy of tasks like recognizing digits or text, such as a vehicle's number plate.