Path: csiph.com!news.mixmin.net!aioe.org!gW6n10AouUGwihugEVAl3g.user.gioia.aioe.org.POSTED!not-for-mail From: Francois LE COAT Newsgroups: sci.image.processing Subject: Temporal Disparity Date: Thu, 14 Jan 2021 18:15:40 +0100 Organization: Aioe.org NNTP Server Lines: 26 Message-ID: NNTP-Posting-Host: gW6n10AouUGwihugEVAl3g.user.gioia.aioe.org Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: quoted-printable X-Complaints-To: abuse@aioe.org User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:60.0) Gecko/20100101 Firefox/60.0 SeaMonkey/2.53.5.1 X-Mozilla-News-Host: news://nntp.aioe.org:119 X-Notice: Filtered by postfilter v. 0.9.2 Xref: csiph.com sci.image.processing:4528 Hi, I recently found that you can compute a "temporal disparity" from two successive images. With an image sequence, performing the perspective registration of the sequence, you can obtain a disparity measure. That means depth with only one camera, though disparity is usually measured by binocular vision. This appears in the following demonstration... A drone is flying between a forest of trees in the French Vosges. Thanks to the optical-flow measured on successive images, the "temporal disparity" reveals the forest of trees... You take a reference image, and the optical-flow is obtained on two rectified images, then the reference is changed when inter-correlation drops below 60%. You can perceive the relief in depth with a unique camera, over time. Happy New Year 2021. Regards, --=20 Dr. Fran=E7ois LE COAT CNRS - Paris - France