基于机器视觉的水面漂浮物智能识别系统

An Intelligent Recognition System for Floating Objects on Water Surface Based on Machine Vision

本文涉及一种基于机器视觉的水面漂浮物智能识别系统,可在江河湖泊的环境中完成对水面漂浮物的识别,属于机器视觉和数字图像处理领域.本文将从水面可疑漂浮物图像、判定图像的采集、初步显著图提取、近景图像预处理、SLIC图像分割、显著目标获得和特征提取、最终利用(back propagation,BP)神经网络对漂浮物性质判定等几个方面进行阐述.

This paper relates to a machine-based intelligent recognition system for surface floating objects,which can identify floating surface objects in the environment of rivers and lakes,and belongs to the field of machine vision and digital image processing.This paper will judge the floating property from the surface suspicious floating object image,the judgment image acquisition,the preliminary saliency map extraction,the close-up image preprocessing,the SLIC image segmentation,the salient target acquisition and feature extraction,and the back propagation(BP)neural network.I will elaborate on several aspects.