News
App analyses tomato appearance within seconds
For researchers and breeders it is important to measure and record the characteristics of a fruit. This is now done manually, which is time-consuming and quite subjective. The Greenhouse Horticulture and Flower Bulbs Business Unit of Wageningen University & Research is developing an app for smartphones that can record the size and shape of a tomato within a few seconds. The app can be used for so-called Distinctness, Uniformity and Stability testing.
By using deep learning, the app has been trained to recognise tomatoes and the circumference of the tomato. This allows the computer vision app to quickly map a number of variables of the tomato, such as the size, volume, weight and the size of the peduncle (the attachment of the stem). The app also provides information about the shape of the tomato.
Using the app is very simple and fast. The app analyses the photographed tomato within a few seconds. By also photographing a QR code in addition to the tomato, the app can also group the photo with associated photos. The app is called MoRPH (Mobile-based Rapid PHenotyping) and is now available via the Google Play Store, but only for closed testing by current project partners. However, it will be publicly available for other interested Android and iOS users by the end of the project duration (middle to end of 2024).
WUR will soon present the app during the International Conference on Computer Vision (in October 2023 in Paris, among others). The app is being developed within the European Horizon2020 project INnovations in plant VarIety Testing in Europe (INVITE). After further development of the app, WUR will also develop the app for image recognition of other fruit vegetables.