Design And Application Of The Labeling Algorithm Of8-adjacent Connecting Area For Massive Gray Scale Images

When the production processing of remote sensing products-namely, the image data, such as thelandscape pattern index products, the urban environmental remote sensing application products, theproducts of the National Nature reserve’s dynamic monitoring and evaluation, the products of nationalecological quality remote sensing monitoring and evaluating, the products of the comprehensive evaluationof national regional ecological construction effect, the products of the global environmental changemonitoring and evaluating, etc-must firstly split the remote sensing image data to achieve segmentationtarget recognition and then realize information extraction in many remote sensing processing production,Since the labeling of connecting area is the necessary step of analysis of the segmentation results. So howremote sensing image data in order to satisfy the demand of smooth boundaries between labeled connectionareas and good image display, is an important issue. To this end, increasing Connection area, such as from4-adjacent connection to8-adjacent connection, seems to be reached the purpose. However, consideringthat the remote sensing data is a kind of gray image and it is massive, calculation will increase significantly,the algorithm efficiency will be in serious decline. Thus, finding a effective labeling algorithm of8-adjacent connecting area for massive gray scale images becomes one of the key issues.To solve the above problem, this paper designs and proposes a labeling algorithm of8-adjacentconnecting area for massive gray scale images, and study how the algorithm is applied to the actualdevelopment of the remote sensing subsystem prototype production system, which is the one of the majorstate issues. The following is the paper work and contributions:(1) Proposed the labeling algorithm of8-adjacent connecting area for massive gray scale images. Forthe problem that existing labeling algorithms of connecting area can’t apply to labeling of8-adjacentconnecting area for the remote sensing image data, based on the analysis on existing connected componentlabeling algorithm, optimizing the traditional labeling algorithm of8-adjacent connecting area by findingfrom four directions of every pixel instead of from eight directions in traditional labeling algorithm,employing the list mechanism to solve labeling conflict of connecting area, and introducing the screeningmechanism and the block processing mechanism, the algorithm realizes fast processing in8-adjacent connecting area labeling for massive gray scale images. At the same time, the algorithm also gets a correctdivision in the connecting area of gray scale images and the accurate number of connecting area.Experimental result shows that the algorithm has high efficiency and small number of labeling conflict, it ismore adaptable to fulfill the requirements of massive gray scale images.(2) The proposed labeling algorithm of8-adjacent connecting area for massive gray scale images isapplied to the actual development of the remote sensing subsystem prototype production system, which isthe one of the major state issues. Combining actual system design and development process of the NationalRemote Sensing Project-remote sensing subsystem prototype production system, this algorithm is appliedto project development effectively. It successfully solves the problem that how the National RemoteSensing Project-remote sensing subsystem prototype production system labels the connection areas of theremote sensing image data, and makes the system completely split the remote sensing image data toachieve segmentation target recognition and realize information extraction, and then makes production ofrelated remote sensing products to be normal, such as the landscape pattern index products, the urbanenvironmental remote sensing application products, the products of the National Nature reserve’s dynamicmonitoring and evaluation, the products of national ecological quality remote sensing monitoring andevaluating, the products of the comprehensive evaluation of national regional ecological construction effect,the products of the global environmental change monitoring and evaluating, etc.