British researchers have developed a machine learning platform that works with ultra-large-scale images taken from the sky and helps monitor lettuce crops in the fields.

A new software called AirSurf-Lettuce measures crop yield, size and location; this will help farmers to harvest with high accuracy, namely from those areas where crops are already fully ripe.

The platform will also help farmers to deliver harvested crops to the market as efficiently as possible. It is important to note that this technology can be applied to other cultures.

The machine was developed at the Earlham Institute (EI) by researchers from the Zhou Group in collaboration with the farm business of G's Growers from Cambridgeshire.

Traditionally, monitoring the state and measuring the amount of future crop yield in the fields is extremely time-consuming and error prone; therefore, new aerial photography based artificial intelligence solutions are a more efficient method.

AirSurf technology uses “deep learning” —a deep structured machine learning technique — combined with ultra-wide image analysis to measure lettuce at high throughput.

This allows you to identify the exact number and location of lettuce plants with the added benefit of recognizing crop quality, that is, small, medium or large lettuce heads.

The combination of this system with GPS allows farmers to accurately track the size distribution of lettuce in the fields, which can help to improve the accuracy and efficiency of agricultural practices, including harvest time.

Lettuce in the UK is big business, especially in East Anglia. Annually, 122,000 tons of it are produced in the country.