Introduction

Linear dependence between variables is the simplest of all possible options. It can be found in many applications, from approximation and geometry tasks, to data compression, camera calibration, and machine learning. But despite its simplicity, things get complicated when real-world influences come into play. All data gathered from sensors includes a portion of noise, which can lead systems of linear equations to have unstable solutions. computer vision problems often require solving systems of linear equations. Even in many OpenCV functions, these linear equations are hidden; it's certain that you will face them in your computer vision applications. The recipes in this chapter will acquaint you with approaches from linear algebra that can be useful and actually are used in computer Vision.