Selective plant detection and treatment using green luminance photometric machine vision scan with real time chromaticity operations and image parameter floors for low processing load
US10269107B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Feb 23, 2017 |
| Grant date | Apr 23, 2019 |
| Priority date | — |
| Expiry date | Dec 9, 2037 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04N23/84
- WIPO fieldAudio-visual technology
- WIPO sectorElectrical engineering
Abstract
A field image is formed using a tristimulus color model and used to detect target plants or entities on a field. Through the use of a luminance parameter floor, hue and saturation selection steps, feature recognition, a sizing floor and an aspect ratio ceiling, a very fast way is devised to recognize a target plant without need for consulting plant attribute databases, or to analyze spectral or other specialized data for comparison with known attributes. This allows a low calculational processing load and use of simple hardware such as a single board computer to handle machine vision in real time. Treatment steps can include a spray event, a light treatment, and a thermal/mechanical trauma.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.