Detection Method of Hydroponic Lettuce Seedlings Status Based on FCN Grid Location and Feature Fusion

In order to find problematic seedlings status timely and improve sorting efficiency of seedlings in cultivation stage of hydroponic vegetables, an automatic detection method of hydroponic lettuce seedlings status based on FCN grid location and feature fusion was proposed, taking dead and double-planting status of seedlings growing in a hole as research object．Aiming at problem of low detection accuracy of two-plant seedlings status, FCN architecture was introduced to change traditional localization based on regression and adopted its sensitive characteristics to obtain accurate grid point spatial information on basis of previous research．At the same time, feature fusion strategy was used to fully obtain correlation between different grid points, so as to achieve further accurate location of problematic status of hydroponic lettuce seedlings．Experimental results showed that mean average precision of this method was 88.1%, which was higher than that of previous method, FSAF, YOLOv3, FoveaBox, ATSS and CornerNet．In particular, detection accuracy of two-plant seedlings status was significantly increased．Hydroponic lettuce seedling condition detection method proposed could realize automatic detection of problem status of hydroponic lettuce seedlings, and provide technical support for intelligent breeding and sorting of hydroponic vegetable seedlings and planting automation．Therefore, the method proposed could realize accurate identification and localization automatically, which could provide technical support for intelligent sorting and automatic planting of hydroponic vegetable seedlings．