Augmented Berry Vision (ABV) - Real-time Augmented Display of Spectral Ripeness Cues in Berry Farms

Demand for healthy products including fresh fruit continues to increase. In the UK soft fruit purchases now account for 22 percent of all consumer fruit purchases. According to Defra's agricultural report, volumes have continued to climb, with the market value of soft fruit grown in the UK worth an estimated at £670 million in 2018\.

There is increasing consumer and retailer demand for high-quality UK-grown soft fruits, and this will increase further post-BREXIT as retailers favour British produce. Great efforts have been made in the past decades to breed varieties with better taste and appearance, whilst investments in improving growing practices has led to production and quality gains. However, achieving consistently ripeness and quality across variable and challenging growing seasons is difficult, and there is a large variability of fruit ripeness after picking due to the invisible changes in colour during ripening. Despite the industry's best efforts in picking, punnets often still contain under or over ripe fruit, leading to negative impact on the perception of fruit quality by consumers.

We will develop a low-cost, spectral imaging based Augmented Reality (AR) prototype glasses device for pickers that can determine and label the fruit ripeness in real-time and assist growers to produce high quality berries with consistent ripeness in every punnet, leading to reduced waste throughout the value-chain. As a proof of concept, blackberry will be used as an exemplar.

The blackberry market in the UK is growing by around 20% year-on-year currently reaching £37.4m in annual retail sales, but determination of ripeness by pickers can be extremely difficult due to the very subtle colour differences between almost-ripe, ripe, and over-ripe fruit. In this project, lab-based fruit quality assessments will be conducted to understand their relationship with ripeness, and hyperspectral imaging will be applied to identify a simple spectral index determining the ripeness.

A lightweight AR frame and LED display will be custom-built to satisfy the application and cost requirement. By integrating a spectral filter or fast tuning narrow-band LED light source, spectral imaging will be processed in real time, and the ripeness of individual berries labelled on the LED display.

The AR technology based picking glasses can be extended to other soft fruits and will have a large impact on the UK soft fruit industry, which as a project partner will have first access to the technology, making a step-change in consistent fruit quality, competitiveness and so economic impact.