Section 6.6 Glossary

Machine Learning (ML)

🔗

An approach to building robot behavior by learning patterns from data, rather than hand-coding every rule explicitly.

🔗

Supervised Learning

🔗

A machine learning approach that trains a model on labeled feature-label pairs (\(y = f(X)\)) to predict outputs for new inputs.

🔗

Unsupervised Learning

🔗

A machine learning approach that works from features (\(X\)) alone, with no ground-truth labels, discovering hidden patterns or clusters (e.g., \(k\)-Means Clustering) based on statistical similarity.

🔗

Dataset

🔗

A collection of examples, each pairing a set of features with its correct label, fed to a machine learning algorithm.

🔗

Feature

🔗

A measurable input value, such as a pixel color or coordinate, used by a model to make predictions.

🔗

Label

🔗

The correct output associated with a training example, such as a category name.

🔗

Classification

🔗

A supervised learning task that predicts a discrete category for a given input.

🔗

Regression

🔗

A supervised learning task that predicts a continuous numeric value for a given input.

🔗

Decision Tree

🔗

A classification or regression model that makes predictions by following a sequence of feature-based decision rules.

🔗

Pixel

🔗

A single picture element in a digital image’s 2D grid, the smallest unit a camera image is broken into.

🔗

Color Channel (RGB)

🔗

One of three numerical values (Red, Green, Blue), each ranging from 0 to 255, that together describe a pixel’s color.

🔗

HSV (Hue, Saturation, Value)

🔗

A color space that separates a pixel’s pure color type (Hue) from how vivid it is (Saturation) and how bright or dark it is (Value), making it more resilient to lighting changes than RGB.

🔗

Color Thresholding

🔗

An image processing technique that classifies each pixel as matching or not matching a target color range.

🔗

Binary Mask

🔗

The output of color thresholding: an image where pixels inside the target color range are set to 1 (white) and all others are set to 0 (black).

🔗

Bounding Box

🔗

The smallest rectangle that contains a detected region of interest in an image, defined by a top-left corner \((x_{\min}, y_{\min})\) and bottom-right corner \((x_{\max}, y_{\max})\text{,}\) from which width, height, and center coordinates \((C_x, C_y)\) are computed.

🔗

Visual Servoing

🔗

Using visual feedback, such as a bounding box’s size and center position, to guide a robot’s steering and approach toward a target in real time.

🔗