Research & Realization Rape Image Extracting System Based On Embedded Linux

It is judged experts methods to monitor crop growth,crop indentification,phenotypic trait detection(etc.).Traditional method of empirical discrimination is bad conducive to accurately the subsequent growth parameters of plants,field phenotyping,and automatic crop identification.It also consumes labor and artificial time.Therefore,this paper applies ARM Cortex-A8 processor as a hardware platform and builds up Linux operating system,selecting development board of S5PV210,targeting for oilseed rapes,to research and design a series embedded rapes acquisition,display and extraction system.It will utilize the incorporation of embedded technology and image processing skill to extract rapes automatically more speedy and accurately.The essential objectives of this research are listed as following:Selecting an appropriate soft and hardware platform,researching S5PV210 peripheral hardware circuit,such as serial port,network card,USB cable and so on.Building cross-compilation environment,seting up tftp & nfs server.At the same time,transplanting u-boot and Linux kernel,as well as making a root file system(ext2 format),shifting to development board.Basing on Linux operation system,projecting the images which are gathered by USB camera driver processor.Besides.Transplanting picture decoding library of libjpeg,LCD display driver will decode the rapes pictures(jpg.),transfering adaptive threshold segmentation functions and utilize OpenCV to achieve images' extract function.Exploring traditional Otsu partitioning algorithm,using Matlab2012 b to emulate the data of four different colors.Compared adaptive threshold segmentation and different color index threshold segmentation effects.selecting an appropriate rapes extracting method.