Last week, i was  in mess processing a great deal of datasets across 3 main sources: RS picture, camera data and the OD table, whilst studied deeper about the 4 steps method in traffic  programming. In this window, recent results of 3 various kinds of data processed will be illustrated.

 

first of all, i was struggling with RS images severely for quite a long term. For i chose Un-supervised Classification method, to eliminate and filter some error or unuseful spots remain a big problem.

Clump function presented error spot under 16 pixels, not so accurate as i imagined:

 

secondly, analysis the OD table based on 25 zones in ChangYi.

 

Last, using camera data to abstract traffic flow: 

 

i have told what my main work was during last several days. Data collection and preparation play a key roles in traffic programming project, the same as in my assignments; therefore, i should better futher advance the quality and amount of useful datasets.