Imaging Characteristics Of Ship Wake Bubble Flim Based On Atmospheric Transmission

When the ships,naval vessels and other objects sailing in the water,the wave will be appeared at the stern.This wave is a layered sea layer,which is called ship wake.The phenomenon in reality illustrates that the ship will form a bubble belt at the stern of the ship because of the rotation of the propeller,the impact between waves,air entrainment and so on,which contains a large number of bubbles with different sizes.The bubbles with larger diameter will float rapidly because of the pressure of the sea water and the gas diffusion in the bubbles,and eventually burst.And the bubbles with smaller diameter will be kept for a long time.The research of ship wake depends on the data of bubble size,bubble density and ship speed.In order to infer the draught,course,speed and other relevant information of the ship during the sail,we must master the optical characteristics of the ship wake,as well as the knowledge of heat,electromagnetism,acoustics and other related disciplines.The ship’s draught,course,speed and other related information can be obtained by analyzing the ship’s aerodynamic characteristics,thermodynamic characteristics and electromagnetic characteristics.This project is based on the atmospheric transmission characteristics of the ship wake bubble curtain imaging characteristics.1)This thesis mainly introduces the generating principle and physical characteristics of the bubble curtain,including the size,the shape and the force analysis and even the basic method of measuring the light scattering temperature and the optical characteristics of the bubble.The characteristics of laser propagation in the atmosphere or water are deeply explained,and the corresponding model is constructed.The scattering effect of atmosphere on laser and the scattering effect of single bubble and bubble group in water are analyzed in detail.2)A wake area extraction method based on shape factor filtering is proposed,which uses the basic image processing methods,such as filtering denoising,corrosion expansion,edge segmentation and morphological processing,which is aiming at the diffusion angle information of the wake with large field of view,for the macro wake image data with multiple interference areas,and for the reason of measuring the variation of the diffusion angle of the wake area with high reliability.The image region segmentation method is used to extract the wake area,and the wake contour is fitted by a straight line.This method realizes the high reliability extraction of the macro wake area.The diffusion angle of the macro wake is calculated from this method,and the measured value is basically consistent with the evaluation of the macro wake diffusion angle in the literature at home and abroad.Through the analysis of the processing results,the navigation direction of the ship is estimated.3)In this thesis,a laser slicing method for wake measurement is proposed.A group of continuous wake sections are collected by laser slicing to obtain the spatial distribution of wake bubbles.The MODTRAN software is used to simulate the atmospheric conditions to get the output power of the laser,and the laser slice experiment is carried out to get the image.After that,based on the KNN nearest neighbor learning wake bubble particle recognition method which uses the feature vector for recognition,the bubble particles in various imaging states are classified to construct the feature vectors of different types of bubble particles,and the KNN nearest neighbor learning algorithm is used to form the image.According to the recognition rules of wake slice sequence,the wake slice images are recognized and counted by the recognition rules.This method can successfully identify most of the bubble particles in the wake sequence images in superposition,occlusion,incomplete imaging and other imaging states.On the basis of this method,the bubble particles with good imaging are selected to calculate the size information,and the volume density of bubble particles in the slice wake bubble curtain is calculated.4)For the level of the 20μm bubble particles,this thesis proposes a recognition method based on K-mean clustering.First of all,the velocity vector field of wake section image is calculated,and then the velocity vector is clustered based on K-mean clustering algorithm.At last,the centroid of clustering is used as the recognition basis.Small bubble particles with similar motion characteristics in wake slice can be identified,and the average velocity of bubble particles can be calculated by using this method.This paper ultimately gives a research on the influence of wake bubble curtain on the physical characteristics of the vehicle,the geometric characteristics of the ship’s wake field,the influence of ship speed on the wake bubble depth,and the influence of ship tonnage on the wake bubble depth.The influence of these factors on the wake bubble curtain is simulated and the corresponding conclusions are obtained.