A hybrid sensory evaluation system for vegetable breeding selections

The quality of a new vegetable cultivar depends not only on production but also sensory quality, which relative to its established varieties and human sense. Due to no IT system supporting and the large amount of samples, the breeders sometimes make their selections in the breeding farm field based on the feeling and experiences of themselves. In this paper, a computer supported vegetable breeding selections system-AgriSense is designed. The vegetable phenotype attributes are stored in a database, the objective measured data of the vegetable samples are gathered, a series of evaluation models are built based on vegetable sensory profile and their related evaluation computing methods, which provide simple interfaces for the breeders to gather data and finish the sensory evaluation step by step. Hybridizing the objective data and the historical data with sensory evaluation, a comprehensive analysis for better breeding selections decision support can be gained.