1 / 41

Improving Classification Accuracy Using Knowledge Based Approach

Title. Improving Classification Accuracy Using Knowledge Based Approach. Ali A. Alesheikh A. Talebzadeh F. Sadeghi Naeeni. Content. Image interpretation by computer vision Traditional strategies Knowledge-based Levels of processing and representation

brie
Download Presentation

Improving Classification Accuracy Using Knowledge Based Approach

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Title Improving Classification Accuracy Using Knowledge Based Approach Ali A. Alesheikh A. Talebzadeh F. Sadeghi Naeeni

  2. Content • Image interpretation by computer vision Traditional strategies Knowledge-based • Levels of processing and representation • Theory and concepts of knowledge-based system Various errors in remotely sensed image analysis Techniques for knowledge representation • use of external knowledge for image interpretation Use of prior probabilities in the decision rule Use of other images as external knowledge • Implementation

  3. Image Analysis Image Processing information Image Artificial Intelligence Computer Graphics Image Interpretation by Computer Vision : - Traditional strategies use spectral information in the image use very little knowledge about the domain the most commonly used approaches in RS have various problems - Knowledge-based image interpretation tends to use more external information in the inference process

  4. Levels of Processing and Representation Levels of processing Levels of representation GIS Knowledge base High (LTM) Matching goal achievement inference high Symbolic description Hypothesis database Intermediate (STM) Segmentation Feature extraction Pre-processing low Image data low

  5. Why Knowledge-basde Image Interpretation? - Various errors in remotely sensed image analysis • During data acquisition process

  6. Why Knowledge-basde Image Interpretation? - Various errors in remotely sensed image analysis • During data acquisition process • Nature of data Adjacent pixels have influence on each other

  7. Why Knowledge-basde Image Interpretation? - Various errors in remotely sensed image analysis • During data acquisition process • Nature of data Adjacent pixels have influence on each other Land cover types do not fit into multiples of rectangular spatial units

  8. Why Knowledge-basde Image Interpretation? - Various errors in remotely sensed image analysis • During data acquisition process • Nature of data Adjacent pixels have influence on each other Land cover types do not fit into multiples of rectangular spatial units Different surface materials may be distinguished by very subtle differences in their spectral patterns

  9. Why Knowledge-basde Image Interpretation? - Various errors in remotely sensed image analysis • During data acquisition process • Nature of data Adjacent pixels have influence on each other Land cover types do not fit into multiples of rectangular spatial units Different surface materials may be distinguished by very subtle differences in their spectral patterns • During classification process

  10. inheritable object non-inheritable declarative optional Knowledge -a priori domain dependent relational essential negative heuristic procedural algorithm Types of knowledge

  11. Techniques for knowledge representation • production rules • semantic network IF <condition> THEN <action> represents objects and relations between objects as a graph structure i.e. a set of nodes connected by labeled arcs • frames or schemas In a frame-based system the objects at each node in the network is defined by a collection of attributed, slots, and values of those attributes, called fillers. Each slot can have procedures attached to it

  12. Example of each knowledge representation techniques Rule #1 IF a pixel feature is (92,99,91) THEN it is “W (Wheat)”or“BID (Barely)” or“SB (Sugar beet)” or“ALO (Alfalfa)”. Rule #2 IF a region in Aster's NDVI map is lower than 0.15 e THEN it's crop type will be W (Wheat) or BID (barely). Rule #3 IF last year's crop was MS THEN in the interest year the crop will be W (Wheat).

  13. MG Maximum probability in traditional classification (e.g. maximum likelihood classification) is a Last year's crop was MS is a is a is a is a is a is a MS MF ALO BID SB W Value on Aster's NDVI map on August <0.15 Example of each knowledge representation techniques

  14. Example of each knowledge representation techniques Frame “W,BID,SB,ALO” slot: they are: W(Wheat),BID(Barely),SB(Sugar beet),ALO(Alfalfa). procedure: if identification of them is desired then search pixels that have maximum probability in any traditional classification like maximum likelihood classification. End frame Frame “W,BID” slots : they are: W(Wheat), BID(Barely). criterion for reconnaissance: they are harvested on the middle of June. procedures: if recognition of W or BID between recognized W, BID, SB, ALO is desired then search areas on Aster's NDVI map which is lower than 0.15. End frame Frame “W” slots : is: W(Wheat), is generalization of: W17, W22, WAT, WTN, WP, WKU, WGP. criterion for reconnaissance: for using the soil in the best way to producing crops, crop calendar disciplines must be considered. procedures : if reconnaissance of W between recognized W, BID is desired then we can use crop calendar disciplines, e.g. search the areas that their last year's crop was MS(Maize Seed). End frame

  15. Probability Real distribution of class 1 P{wk ,Xi } = F k(Xi) P{wk } ^ (-1) (X-m) Fk=(2p)-p/2|Sk|-1/2 e-1/2(X-m)'S A posteriori probability of class 2 given equal a prior probability Fk(Xi ) P{w k , v j } P{w k,Xi } = P {w k| Xi} P{w k | Xi , v j} = A posteriori probability of class 1 given equal A prior probability P{X} SKk=1 Fk(Xi ) P{w k , v j } Real distribution of class 2 Feature Estimated threshold Real threshold using of external knowledge for image interpretation - Using of prior probability in the decision rule (maximum likelihood approach)

  16. P{wk ,Xi } = F k(Xi) P{wk } ^ (-1) (X-m) Fk=(2p)-p/2|Sk|-1/2 e-1/2(X-m)'S Fk(Xi ) P{w k , v j } P{w k,Xi } = P {w k| Xi} P{w k | Xi , v j} = P{X} SKk=1 Fk(Xi ) P{w k , v j } using of external knowledge for image interpretation - Using of prior probability in the decision rule (maximum likelihood approach) - Using of other images as external knowledge The other knowledge for interpretation can be the other image which is acquired in the other time or with the other sensor. The resolution and spectral bands of the other image can be different from initial one.

  17. Implementation Study area • Moghan plain located in Ardebil

  18. Implementation Study area • Moghan plain located in Ardebil • About 300,000 tons of various crops produce annually in 18000 ha of irrigated farms.

  19. Implementation Study area Corp Acreage(ha) Yield Wheat 7000 up to 6500 kg/ha Barely 1500-2000 up to 5000 kg/ha Sugar Beet 3000 more than 50tons/ha Maize Seed 15000 more than 2500 kg/ha Maize Grain 1500 more than 6500kg/ha Maize Silage 800 more than 40tons/ha Alfalfa 1500 about 12tons/ha Forage crops 700 20-100tons/ha • Moghan plain located in Ardebil • About 300,000 tons of various crops produce annually in 18000 ha of irrigated farms.

  20. Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid)

  21. Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid) • Map of field boundaries ( production of polygonized fields)

  22. Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid) • Map of field boundaries ( production of polygonized fields) • Data about crop type of each field

  23. GIS of Moghan Fields Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid) • Map of field boundaries ( production of polygonized fields) • Data about crop type of each field • ETM+ image (color composite 354) (was acquired on 2001-05-23)

  24. GIS of Moghan Fields Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid) • Map of field boundaries ( production of polygonized fields) • Data about crop type of each field • ETM+ image (color composite 354) (was acquired on2001-05-23) • Aster image (was acquired on August 2001-8-23)

  25. GIS of Moghan Fields Georeferenced by map on 1/50000 scale Available DATA: • Maps of study area in 1/50000 scale (UTM coordinate system and in WGS84 ellipsoid) • Map of field boundaries ( production of polygonized fields) • Data about crop type of each field • ETM+ image (color composite 354) (was acquired on 2001-05-23) • Aster image (was acquired on August 2001-08-23)

  26. Experimental work • Spectral-based : • Crop rotation patterns • Times of planting and harvesting • Field boundaries information • Climate information • Geographical information

  27. Experimental work • Knowledge-based : • Spectral-based : • Crop rotation patterns • Times of planting and harvesting • Field boundaries information • Climate information • Geographical information

  28. Experimental work • Knowledge-based : • Spectral-based : • Crop rotation patterns • Times of planting and harvesting • Field boundaries information • Climate information • Geographical information

  29. Experimental work • Knowledge-based : • Spectral-based : • Crop rotation patterns • Times of planting and harvesting • Field boundaries information • Climate information • Geographical information • Financial information • Crop 'portfolio management' • Agricultural information • Advice centers

  30. Spectral-based : rule matrices of every seven crop based on maximum likelihood approach and equal prior probability

  31. Spectral-based : rule matrices of every seven crop based on maximum likelihood approach and equal prior probability

  32. Spectral-based : rule matrices of every seven crop based on maximum likelihood approach and equal prior probability Overall accuracy of spectral-based classification = 53.2%.

  33. Spectral-based : rule matrices of every seven crop based on maximum likelihood approach and equal prior probability Overall accuracy of spectral-based classification = 53.2%.

  34. Knowledge-based classification : - Using of Crop Rotation Patterns : • Transition matrix production TRANSITION MATRIX "1998-1999, 1999-2000" ALO BID MF MG MS SB W ALO 0.86 0 0 0 0 0 0.14 BID 0 0.33 0.21 0.2 0.07 0.19 0 MF 0 0.49 0 0 0.04 0 0.47 MG 0 0.18 0.01 0 0 0 0.81 MS 0 0 0 0 0 0 1 SB 0 0.02 0 0 0.03 0 0.95 W 0 0.17 0.05 0.03 0.43 0.23 0.09 TRANSITION MATRIX "1997-1998, 1998-1999" ALO BID MF MG MS SB W ALO 0.91 0 0 0 0 0 0.09 BID 0 0.38 0.22 0.14 0.09 0.17 0 MF 0 0.47 0 0 0 0.05 0.48 MG 0 0.09 0 0 0 0 0.91 MS 0 0 0 0 0 0 1 SB 0 0 0 0 0 0 1 W 0 0.17 0 0.07 0.32 0.38 0.06

  35. Knowledge-based classification : - Using of Crop Rotation Patterns : Stable Dynamic System • Comparison between them TRANSITION MATRIX "1998-1999, 1999-2000" ALO BID MF MG MS SB W ALO 0.86 0 0 0 0 0 0.14 BID 0 0.33 0.21 0.2 0.07 0.19 0 MF 0 0.49 0 0 0.04 0 0.47 MG 0 0.18 0.01 0 0 0 0.81 MS 0 0 0 0 0 0 1 SB 0 0.02 0 0 0.03 0 0.95 W 0 0.17 0.05 0.03 0.43 0.23 0.09 TRANSITION MATRIX "1997-1998, 1998-1999" ALO BID MF MG MS SB W ALO 0.91 0 0 0 0 0 0.09 BID 0 0.38 0.22 0.14 0.09 0.17 0 MF 0 0.47 0 0 0 0.05 0.48 MG 0 0.09 0 0 0 0 0.91 MS 0 0 0 0 0 0 1 SB 0 0 0 0 0 0 1 W 0 0.17 0 0.07 0.32 0.38 0.06

  36. Terrain object data (t-1) IF last year's crop = Wheat THEN current crop = Barely (17%), Maize feed (5%), Maize grain (3%), Maize seed (43%), Sugar beet (23%), Wheat (9%). Remote sensing data (t) Information extraction Updating GIS Application context Overall accuracy of maximum likelihood and estimated prior probability 66.7%.

  37. Knowledge-based classification : > 0.15 < 0.15 - Times of planting and harvesting Wheat and Barely are harvested on the June Using of NDVI produced from Aster image which was acquired on 23 August 2001

  38. Knowledge-based classification : - Times of planting and harvesting Wheat and Barely are harvested on the June Using of NDVI produced from Aster image which was acquired on 23 August 2001 IF value of NDVI map is smaller than 0.15 THEN crop type will be W(Wheat) or BID(barely) IF produced probability of W from the previous step is greater than probability of BID THEN crop type will be W(Wheat) Overall accuracy of knowledge-based classification = 72.3 %.

  39. Knowledge-based classification : - Field boundaries information In each field one crop type Overall accuracy of knowledge-based classification = 88.7 %.

  40. conclusion This paper shows us that "traditional image analysis seems to be like a random walk in problem space" and by using any external knowledge, known way can be selected for receiving the goal.

  41. Future works • Crop rotation was used in this thesis. Transition matrices were produced from two successive years. They can be extracted from three, four or more successive years. • Other data sources can be used as external knowledge, e.g. the otherbands of aster image can help us for interpretation. • Knowledge about local soil types and conditions could be used to help predict likely crops to be planted. • We can use geographical information as an external knowledge. E.g. economical constraints affect likelihood of crops. For example, crops with a high transportation cost and low profit margin may become less probable the further away from a storage silo the field is. • Financial information can help us for image interpretation. By this fact that, farmers also base their decisions about which crops to plant based on market potentials, aiming to maximize profitability. Information about expected crop prices and likely future demand could again assist in classification

More Related