Census transform

<p>The census transform (CT) is an image operator that associates to each <a href="/facts/Pixel/UzwBvXBV">pixel</a> of a <a href="/facts/Grayscale/Tjb9kR8v">grayscale</a> image a binary string, encoding whether the pixel has smaller intensity than each of its neighbours, one for each bit. It is a non-parametric transform that depends only on relative ordering of intensities, and not on the actual values of intensity, making it invariant with respect to <a href="/facts/Monotonic_function/MjQK0YL2">monotonic</a> variations of <a href="/facts/Illumination_(image)/rxqFN6Dm">illumination</a>, and it behaves well in presence of multimodal distributions of intensity, e.g. along object boundaries. It has applications in <a href="/facts/Computer_vision/Tl2Yyk66">computer vision</a>, and it is commonly used in visual <a href="/facts/Correspondence_problem/YBe0HbVX">correspondence problems</a> such as <a href="/facts/Optical_flow/ju6d82wz">optical flow</a> calculation and <a href="/facts/Binocular_disparity/Fj2I1Soj">disparity</a> estimation.
</p><p>The census transform is related to the rank transform, that associates to each pixel the number of neighbouring pixels with higher intensity than the pixel itself, and was introduced in the same paper.
</p>

Census transform open-in-new

Census transform