Development of a smartphone-based peanut data logging system

The southeastern states of the USA, namely Georgia, Florida and Alabama, account for two-thirds of the total peanut production in the USA. Determining the optimal harvest day for peanuts is critical because it directly impacts their yield and grade. The conventional method is to identify the percentage of black peanuts in representative field samples using a peanut profile board. In this study, a new assistive system for peanut data logging was developed, which was able to upload user information and harvest conditions to an online database. Field tests showed the system had errors of 9.0%, 18.8% and 5.8% in estimating the percentage of black-and-brown peanuts, the percentage of black peanuts and the total peanut number when compared to results from human graders, respectively. This proof-of-concept data logging system has the potential to provide fast and reliable results. Future work will include applying a segmentation algorithm to increase the accuracy of the system as well as creating a harvest prediction model using the online database.