Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. 2011 IEEE International Geoscience and Remote Sensing Symposium<br />26/07/2011<br />Vancouver, Canada<br />SATELLITE IMAGE ARTIFACTS DETECTION BASED ON COMPLEXITY DISTORTION THEORY<br />Avid ROMAN GONZALEZ<br />Mihai DATCU<br />Avid Roman Gonzalez<br />
  2. 2. OUTLINE<br />The Artifacts, problematic.<br />Rate-Distortion Function<br />Kolmogorov Complexity<br />Kolmogorov’s Structure Function<br />Experiments and Results<br />Conclusions<br />Avid Roman Gonzalez<br />
  3. 3. ARTIFACTS :<br />The artifacts are artificial structures that represent a structured perturbation of the signal. Therefore, these artifacts induce errors in the indexation of the images.<br />3/47<br />Avid Roman Gonzalez<br />
  4. 4. ARTIFACTS KINDS :<br />Aliasing<br />Strips<br />. . .<br />Saturation<br />Blocking<br />4/47<br />Avid Roman Gonzalez<br />
  5. 5. DATA CLEANING:<br />Data cleaning, or data cleansing or scrubbing<br /><ul><li> Detecting and removing errors and inconsistencies from data in order to improve the quality of data .</li></ul>Data qualityproblems are present in single data collections, such as files and databases, e.g.:<br />- Due to misspellings during data entry.<br />- Missing information or other invalid data.<br />5/47<br />Avid Roman Gonzalez<br />
  6. 6. CLASICAL APPROACH :<br />Is to predict or determine the existence of defects, to model it, and then design a method to detect and correct them. For example we have the lines correction methods presented by [Hyung-Sup Jung 2009].<br />SpecificArtifacts.<br />OUR APPROACH :<br /><ul><li>Using data compression techniques to implement a method more generic PARAMETER FREE regardlessthe type or model of artifact.</li></ul>Hyung-Sup Jung, Joong-Sun Won, Myung-Ho Kang, and Yong-Woong Lee, “Detection and Restoration of Defective Lines in the SPOT 4 SWIR Band”, Transaction on Image Processing, 2009.<br />
  7. 7. RATE-DISTORTION FUNCTION:<br />The Rate-Distortion (RD) Functionis given by the minimum value of mutual information between source and receiver under some distortion restrictions.<br />The RD function shows how much compression (lossy compression) can be used without loss of distortion preset value.<br />7/47<br />Avid Roman Gonzalez<br />
  8. 8. Images with different <br />compression factor (cf)<br />Imagecf 1<br />Imagecf 2<br />Imagecf 2<br />.<br />.<br />.<br />Imagecf n<br />Decompression<br />Image<br />Features Vector<br />(compressionerrors)<br />JPEG<br />LossyCompression<br />+-<br />Classification<br />For the artifacts detection, we propose to use the RD function obtained by compression of the image with different compression factors and examine how an artifact can have a high degree of regularity or irregularity for compression.<br />8/47<br />Avid Roman Gonzalez<br />
  9. 9. KolmogorovComplexity<br />is the length of a shortest program to compute x on a universal Turing machine<br />K(x) is a non calculable function<br />15 x (Write 0)<br />Write 1010001010111011<br />9/47<br />Avid Roman Gonzalez<br />
  10. 10. Kolmogorov’sStructureFunction<br />An approximation of the RD curve using the Kolmogorov complexity theory could be the Kolmogorov Structure Function (KSF).<br />The relation between the individual data and its explanation (model) is expressed by Kolmogorov’s structure function.<br />The original Kolmogorov structure function for a data x is defined by:<br />Where: S is a contemplated model for x.<br />αis a non-negative integer value bounding the complexity of the contemplated S.<br />10/47<br />Avid Roman Gonzalez<br />
  11. 11. Kolmogorov’sStructureFunctionforTextureDiscrimination<br />To evaluate the behavior of the KSF for different textures, we use de Brodatz images databases. We show the textures used for this experiments.<br />Avid Roman Gonzalez<br />
  12. 12. We can observe that the KSF can discriminate more or less the different structure, the curve KSF has a similar shape for each texture group, but the level is different.<br />Avid Roman Gonzalez<br />
  13. 13. ArtifactsDetectionUsingKolmogorov’sStructureFunctionApproach<br />To detect artifacts using Kolmogorov Structure Function (KSF), the first step is to watch the behavior of the KSF curve for images with artifacts and images without artifact.<br />One aspect to consider is how to generate the candidates for the necessary space S. For this purpose, we have generated the candidates using 2 methods: Candidates generation by JPEG lossy compression and Candidates generation by genetic algorithm.<br />Avid Roman Gonzalez<br />
  14. 14. KSF using jpeg lossy compression<br />KSF using genetic algorithm<br />We can observe that the better discrimination is done when we generate the candidates for the space S using the JPEG lossy compression. Also using JPEG lossy compression the approximation to the Rate-distortion analysis is better.<br />Avid Roman Gonzalez<br />
  15. 15. We use the jpeg lossy compression for generate candidates and to draw the Kolmogorov Structure Function for each patch of a satellite image and try to detect the artifacts. For this experiments we use an image with aliasing introduces manually.<br />Aliasing detection in city environmental using KSF and candidate generation with JPEG lossy compression<br />Avid Roman Gonzalez<br />
  16. 16. CONCLUSIONS<br />The Kolmogorov structure function represents the relationship between an element or data with its model, structure, or explanation.<br />In this work, we have used the Kolmogorov structure function as a approximation of rate-distortion function using Kolmogorov complexity theory and the complexity-distortion theory, so we can examine the complexity of the images to be analyzed, this complexity would be related to the presence or absence of artifacts.<br />16/47<br />Avid Roman Gonzalez<br />
  17. 17. CONCLUSIONS<br />The generation of candidates for to calculation the Kolmogorov structure function is an important step, in this work was done experiments using 2 methods, generation of candidates by jpeg lossy compression and candidate generation using genetic algorithms, we obtain better results using lossy jpeg compression.<br />17/47<br />Avid Roman Gonzalez<br />
  18. 18. THANK YOU<br /> FOR YOUR ATTENTION<br /><br /><br />Avid Roman Gonzalez<br />