This paper introduces author's work in compress to vector spatial data

which will be transported in Internet/Intranet. On the basis of vector spatial data compression method

which is proposed by[1]

the authors proposes next four innovation:(1)For vector spatial data

not only integer but also short data can be used to replace traditional double or float data. (2)The map data are mapped to a“virtual screen”

the formula of“virtual screen”width is: width=DispWidthPixel×MaxEnlargeRatio/stepWidth①. (3) In the mapping process

execute “filter points compression”to pattern points of arc(including polygon' s boundary arc and line' s arc). In the formula①

if stepWidth=1

“filter compression”has no loss in precise of map displayed

else stepWidth>1

it has precise loss in map displayed. (4)“Long edge adding point”to deal with exception of“Offset compress”to pattern points in arc

which offset in adjacent two points may larger than 127. After these compressions

a map vector data may be compressed to 20% or less.