Research Segmentation Methods Of Stumpage Image Based On Android System

The purpose of the study is to develop a system for calculating stumpage parameters that is height and DBH of stumpage based on Android. The method improves the traditional measurement methods about stumpage parameter information and technology.Based on the plant standing image information analysis and processing, segmenting fiduciary object from the stumpage image, image segmentation results will be combined with photogrammetry, and calculate standing height and DBH in the stumpage image.Standing image segmentation algorithm proposed in this paper, based on the RGB color space, using color feature regions, spatial clustering method and color information of standing image to segment standing image. The proposed algorithm will divided the image into two parts of trunk and crown.Standing trunk segmentation algorithm based on region method to choose an area of interest on the standing tree trunk, use algorithm to weed out outliers from the region of interest, according to the region of interest statistics optimized threshold range and RGB color components R, G, Section B of the variance. Based spatial clustering method and analysis of the image color features standing trunk, and add in the initial image segmentation point in line with the trunk part of the color features, while clearly not delete trunk color feature points.Stumpage crown segmentation algorithm based on spatial clustering method. Firstly, to segment color stumpage image based on color features. Then according to the color characteristics of the image pixel, to remove some pixel that isn’t belonged to the crown.In the segmentation result image still retains some pixels which are not the part of trunk or crown, therefore the image segmentation results can not be used to calculate the parameters of standing directly. To get the details as much as possible while to remove more noise interference from the standing image, the paper will divide the stumpage segmentation result image from two parts, and to optimize them by different image processing algorithms. Image segmentation optimization process mainly by 3* 3 templates two-dimensional mean filter and 2*2 templates morphological filtering and 3*3 templates morphological filtering and some algorithms such as removing small area. In this paper, use the algorithm which is similar to the stumpage crown segmentation algorithm to segment the measuring ruler.The proposed algorithm is achieved on Matlab platform. Contrast with the result image based on the gray threshold segmentation, it proves that image segmentation processing algorithms retain more target color characteristics information in the canopy and canopy standing image pixels, while eliminating the some shaded area. It can be to compare the standing part of the complete division.Comparing with the results form this paper proposed segmentation algorithm and the actual measurement proves that relative error is small.In this paper, by standing image segmentation algorithm based on RGB color space to achieve the stumpage image parameter estimation software on the Android mobile platform. The main features of the software is that image acquisition standing, standing image storage, image processing standing, standing parameter estimates, and standing parameter information uploads.