Canonical: https://knowlpedia.com/convex-analysis/image-and-kernel-linear-isomorphism/

# Image, kernel, and linear isomorphism

The image and kernel of a linear operator; bijective linear maps are isomorphisms

Let \(T:X\to Y\) be a [linear operator](https://knowlpedia.com/convex-analysis/linear-operator-linear-transformation/) between vector spaces.

- The **image** of \(T\) is
  \[
  \operatorname{im}T:=T(X)=\{T(x)\mid x\in X\}\subset Y.
  \]

- The **kernel** of \(T\) is
  \[
  \ker T:=T^{-1}(\{0\})=\{x\in X\mid T(x)=0\}\subset X.
  \]

If \(T\) is bijective (one-to-one and onto), then \(T\) is a **linear isomorphism**, and we say that \(X\) and \(Y\) are **isomorphic**, written \(X\cong Y\).

## Remarks

Both \(\ker T\) and \(\operatorname{im}T\) are [linear subspaces](https://knowlpedia.com/convex-analysis/linear-subspace/) (see [images and preimages of subspaces](https://knowlpedia.com/convex-analysis/images-and-preimages-of-subspaces-under-linear-operators/)). The quotient \(X/\ker T\) is canonically isomorphic to \(\operatorname{im}T\) (see [the isomorphism theorem](https://knowlpedia.com/convex-analysis/isomorphism-theorem-and-dimension-formula-for-linear-operators/)).

## Examples

- If \(T:\mathbb{R}^2\to\mathbb{R}\) is \(T(x,y)=x+y\), then \(\ker T=\{(t,-t):t\in\mathbb{R}\}\) and \(\operatorname{im}T=\mathbb{R}\).
- If \(T=\mathrm{id}_X\), then \(\ker T=\{0\}\) and \(\operatorname{im}T=X\); \(T\) is an isomorphism.
