如果我理解正確的話,你可以使用rgeos
包和相關的空間工具R.
我採取了訣竅來緩衝相交線,然後從該網站生成「差異」多邊形:
http://www.chopshopgeo.com/blog/?p=89
生成示例柵格和覆蓋的多邊形。
vdata <- list(x = 1:nrow(volcano), y = 1:ncol(volcano), z = volcano)
## raw polygon data created using image(vdata); xy <- locator()
xy <- structure(list(x = c(43.4965355534823, 41.7658494766076, 36.2591210501883,
25.560334393145, 13.7602020508178, 18.7949251835441, 29.179041644792,
40.6645037913237, 44.2832110429707, 47.272577903027, 47.5872480988224
), y = c(30.0641086410103, 34.1278207016757, 37.6989616034726,
40.900674136118, 32.7732500147872, 27.4781100569505, 22.5523984682652,
22.7986840476995, 24.5226831037393, 29.3252519027075, 33.8815351222414
)), .Names = c("x", "y"))
## close the polygon
coords <- cbind(xy$x, xy$y)
coords <- rbind(coords, coords[1,])
library(sp)
## create a Spatial polygons object
poly <- SpatialPolygons(list(Polygons(list(Polygon(coords, hole = FALSE)), "1")))
## create a contour line that cuts the polygon at height 171
cl <- contourLines(vdata, levels = 171)
## for ContourLines2SLDF
library(maptools)
clines <- ContourLines2SLDF(cl)
現在,相交線的多邊形,然後稍微緩衝行和差再次與多邊形給予多聚。
library(rgeos)
lpi <- gIntersection(poly, clines)
blpi <- gBuffer(lpi, width = 0.000001)
dpi <- gDifference(poly, blpi)
繪製原始數據,並從空間對象手動提取多邊形的一半。
par(mfrow = c(2,1))
image(vdata)
plot(poly, add = TRUE)
plot(SpatialPolygons(list(Polygons(list([email protected][[1]]@Polygons[[1]]), "1"))),
add = TRUE, col = "lightblue")
image(vdata)
plot(poly, add = TRUE)
cl <- contourLines(vdata, levels = 171)
plot(SpatialPolygons(list(Polygons(list([email protected][[1]]@Polygons[[2]]), "2"))),
add = TRUE, col = "lightgreen")
這適用於這個相當簡單的情況,它可能對您的情況有用。
看看http://cran.r-project.org/web/views/Spatial.html – 2011-04-20 07:17:47
Thanks @ Roman〜我會研究它 – Marco 2011-04-20 07:42:52
你能給我們一些玩具的例子嗎?如果你能告訴我們你正在使用哪些對象/軟件包,這也很好。有多種可能性。 – 2011-04-20 11:27:36