Kansas State University is a world leader in the study of wheat genetics to develop new varieties that can tolerate a wide range of environmental conditions. Field Book, a new mobile application, was developed to modernize plant breeding programs around the world. Building off its success, we’ve developed several additional mobile apps for plant breeding. As a part of this effort, novel image analysis algorithms are being developed to model and extract plant phenotypes. This paper describes a dynamic, real-time algorithm to accurately count seeds using a modest mobile device. There are many directions for future research to enhance the algorithm’s performance and accuracy. The new algorithm can be used to count a wide variety of different types of crop seeds for high-throughput phenotyping.