Abstract: Disclosed are a dynamic route planning method and device for unmanned ship used for feeding. The method includes: acquiring positioning information of a first set of vertices in a water body region, a feeding distance of a feeding device and positioning information of an unmanned ship; establishing a regional model according to the positioning information of the first set of vertices, and obtaining a route planning point according to the regional model and the feeding distance of the feeding device; obtaining a route target point according to the positioning information of the first set of vertices and the route planning point; dividing the route target point into an unreached target point or a reached target point according to the positioning information of the unmanned ship and the route target point; and obtaining a cruise route according to the positioning information of the unmanned ship and the unreached target point.

Abstract: A method and device for detecting water quality based on color recognition, and a storage medium. The method includes: collecting a water quality parameter sample and a RGB color parameter sample; normalizing the water quality parameter sample and the RGB color parameter sample; fitting a non-linear curve formed by RGB-PH values by using a RBF neural network model; establishing a RGB-water quality parameter lightweight database by using the non-linear curve of the RGB-water quality parameter; acquiring a color RGB value after a water sample to be detected reacts with a reagent; and comparing the color RGB value with the database to obtain a water quality detection parameter of the water sample to be detected.