Abstract: A training method for multi-output land cover classification model and a classification method are provided. The training method includes: obtaining a training data; inputting the training data into an initial model based on deep belief nets for training to obtain a multi-output land cover classification model, wherein the initial model includes N level outputs, and the N level outputs include an output set at last network layer and (N?1) level output set at any (N?1) network layers from a first network layer to a penultimate network layer of the initial model; determining a total loss according to losses of the N level outputs; performing a backpropagation based on the total loss to adjust a parameter of the initial model, N being an integer greater than or equal to 2. The gradient is not easy to disappear during backpropagation of the model, which is beneficial to improve classification accuracy.

Abstract: Methods and devices for correcting underwater photon displacement and for depth sounding with a single-photon Lidar are provided. The method includes: acquiring a pointing angle of a photon emitted by the single-photon Lidar, and coordinates of a water-surface photon signal and a water-bottom photon signal returned by the photon emitted by the single-photon Lidar; performing a sea wave fitting according to the water-surface photon signal to determine a sea wave model; determining an intersection of the photon and an air-water interface according to coordinates of any water-bottom photon, the pointing angle and the sea wave model; determining an underwater displacement error of the photon according to the intersection, the sea wave model and the pointing angle; and correcting the coordinates of the water-bottom photon according to the underwater displacement error. The invention performs sea wave modeling through water surface photon signal and determines intersection of the photon and water-air interface.

Abstract: An adaptive filtering method of photon counting Lidar for bathymetry is provided in this invention, which includes steps: step S1: adaptively acquiring parameters of elliptic filtering for water surface photon signals; step S2: determining a relationship between filter parameters and elevation of underwater photon signals, and obtaining parameters of the elliptic filtering for photon signal in water column; and step S3: filtering and fitting the water surface photon signals and the underwater photon signals to acquire continuous bathymetry results.

Abstract: A training method for multi-output land cover classification model and a classification method are provided. The training method includes: obtaining a training data; inputting the training data into an initial model based on deep belief nets for training to obtain a multi-output land cover classification model, wherein the initial model includes N level outputs, and the N level outputs include an output set at last network layer and (N?1) level output set at any (N?1) network layers from a first network layer to a penultimate network layer of the initial model; determining a total loss according to losses of the N level outputs; performing a backpropagation based on the total loss to adjust a parameter of the initial model, N being an integer greater than or equal to 2. The gradient is not easy to disappear during backpropagation of the model, which is beneficial to improve classification accuracy.

Abstract: An adaptive filtering method of photon counting Lidar for bathymetry is provided in this invention, which includes steps: step S1: adaptively acquiring parameters of elliptic filtering for water surface photon signals; step S2: determining a relationship between filter parameters and elevation of underwater photon signals, and obtaining parameters of the elliptic filtering for photon signal in water column; and step S3: filtering and fitting the water surface photon signals and the underwater photon signals to acquire continuous bathymetry results.

Abstract: Methods and devices for correcting underwater photon displacement and for depth sounding with a single-photon Lidar are provided. The method includes: acquiring a pointing angle of a photon emitted by the single-photon Lidar, and coordinates of a water-surface photon signal and a water-bottom photon signal returned by the photon emitted by the single-photon Lidar; performing a sea wave fitting according to the water-surface photon signal to determine a sea wave model; determining an intersection of the photon and an air-water interface according to coordinates of any water-bottom photon, the pointing angle and the sea wave model; determining an underwater displacement error of the photon according to the intersection, the sea wave model and the pointing angle; and correcting the coordinates of the water-bottom photon according to the underwater displacement error. The invention performs sea wave modeling through water surface photon signal and determines intersection of the photon and water-air interface.