Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Abstract.

The use of machine vision (MV) in the domain of precision agriculture is currently expanding; repetitive, tedious tasks, such as the sorting and grading of specialty crops, are continuously being computerized to save producers time and money. Technical advancements in MV have improved the detection, quality assessment and yield estimation processes for apple orchards, mangoes, maize, figs and many other fruits. However, similar methods for vegetable crops have yet to be fully developed. This paper describes the initial design process of a prototype, MV-based yield monitor for several vegetable crops produced by Delfland, Inc. This monitor collects and processes yield data by counting harvested vegetables and classifying them by shape and size. The target vegetables include dry shallots, charlotte onions, Chinese radish, carrots and lettuce. The presented MV algorithm uses a watershed segmentation method and color classification process. Based on the preliminary results, occlusions and inconsistent light conditions are the main limiting factors. Although further enhancements are envisioned for the prototype system, when developed fully, it will have the potential to benefit many producers of small vegetable crops.