Sie sind auf Seite 1von 26

Post Processing Post Processing

Of digitally classified imagery Of digitally classified imagery

Post Processing Post Processing


Topics:

Why

process any further? Filtering

Spatial Pixels

to Polygons

Post Processing Post Processing


Why Process Any Further?

Four main reasons to process further:


1.

Manually Edit the Classification

e.g. steep, deeply shaded slopes classified as water..., subject of a subsequent chapter,

1.

Refine the classification using ancillary data,

1. 2.

Summarize (smooth) the classification, Convert the classified data to vector format.

Why Process Further? Why Process Further?


The vector format has advantages in certain applications:
Raster
Resolution: Disk Storage: Set by pixel size Determined by pixel and image size

Vector
Variable Usually more efficient

Limit of Spatial Accuracy: Spatial Analysis... - possible operations: -computational speed: - topology:

Within 1 pixel

Unlimited

Same Usually Greater No explicit topology

Same Usually Less Explicit topology

Post Processing Post Processing


Pixels to Polygons...

In addition:

Users usually find vector format output easier to interpret and, Other data may already be in vector format, Further processing requires time, money and disk space, May alter class boundaries

There are Disadvantages:

How are pixels converted to polygons...?

Raster and Vector Integration Raster and Vector Integration


Creating Polygons from Pixels...

Polygons based on pixel classification,

Automatic conversion to polygons isnt feasible because of too many tiny polygons,

Filters are used to aggregate pixels into more homogenous groupings, Automatic conversion of the aggregated pixel groupings to polygons.

Post Processing--Spatial Filtering Post Processing--Spatial Filtering

Spatial Filtering is an area operator. It creates new map values as a function of values of existing neighboring pixels, Usually summarizes the existing map using a roving window or kernel, Standard kernel operators:

Maximum, minimum, majority, minority, mean, median, mode, standard deviation, diversity and many more.

Moving Window Concept Moving Window Concept

Projection of 3x3 Kernel The Moving Window (kernel) scans the 3x3 neighborhood of every pixel in the classified image.

Classified Image

Moving Window Concept Moving Window Concept

A value is computed, depending on the type of kernel, from the 9 values in the input file and placed in the corresponding cell of the output file.
Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Moving Window Concept Moving Window Concept

Output File

Example: Mean Kernel Calculation Example: Mean Kernel Calculation

8 8 2 8 2 6

9 3 8 6 2 9

9 11 11 11 9 3 9 11 3 5 5 9 8 6 8 11 4 4 6 9 11 4 4 6
or

1/9 1/9 1/9

1/9 1/9 1/9

1/9 1/9 1/9

(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 =

5.44 5

Round(3 + 5 + 5 + 8 + 6 + 8 + 4 + 4 + 6) / 9 =

Example: Mode Kernel Calculation Example: Mode Kernel Calculation

8 8 2 8 2 6

9 3 8 6 2 9

9 11 11 11 9 3 9 11 3 5 5 9 8 6 8 11 4 4 6 9 11 4 4 6

No Formula

In this example: The mode is initially a 3. Then two 5s set the mode to 5. Several other pairs of numbers are then found but they cant beat the pair of 5s. Output =

Mode Kernel Example Mode Kernel Example

The Original Classification: Note the salt and pepper appearance (high spatial frequency)

Mode Kernel Example Mode Kernel Example

The Mode-Filtered Output: Note the salt and pepper appearance (high spatial frequency) is much less.

Pixels to Polygons Pixels to Polygons

The aggregated pixels can be automatically converted to polygons

Pixels to Polygons Pixels to Polygons

The polygons alone

Pixels to Polygons Pixels to Polygons

The variability within polygons is not lost --the polygons can be displayed with the original classification.

Raster and Vector Raster and Vector


Database Integration:

Can use existing polygons, Summarize pixel class values, Indicates the intra-polygon variation.

Raster and Vector Raster and Vector


Database Integration:

Composition of Polygon 47

47

class 7 8 9

acres 30 60 110

% 15 30 55

description crown closure 10 - 40% crown closure 41 - 70% crown closure 71 -100%

Das könnte Ihnen auch gefallen