mat lab ppt final.pptx for imahe under water

its a matlab project

1. KAKINADA INSTITUTE OF ENGINEERING AND TECHNOLOGY Department of ELECTRONICS and COMMUNICATION engineering UNDERWATER IMAGE ENHANCEMENT USING ADAPTIVE RETINAL MECHANISM PRESENTED BY: T. MANOJ KUMAR REDDY 18B21A04F4 A.GAYATHRI 19B25A0403 K.BALA KRISHNA 18B21A04K4 K.SOMASEKHAR 18B21A0418 N.PRAVEEN PRASAD 18B21A04JO UNDER THE GUIDANES OF: Mr. D DHARMA SASTRA, MTECH ASST PROF, ECE DEPT.

2. CONTENT OVERVIEW Objective Abstract Introduction Proposed method Retinex and noise aware shadow-up Decomposition based on retinal theory WVM AGCWD Shadow-up function YUV BM3D Results Advantages and Applications Conclusion and Future scope

3. ABSTRACT The proposed approach has mainly three steps to solve the problems mentioned above. First, a simple but effective colour correction strategy is adopted to address the colour distortion. Second, a variational framework for retinex is proposed to decompose the reflectance and the illumination, which represent the detail and brightness respectively, from single underwater image. An effective alternating direction optimization strategy is adopted to solve the proposed model. Third, the reflectance and the illumination are enhanced by different strategies to address the under-exposure and fuzz problem. The final enhanced image is obtained by combining use the enhanced reflectance and illumination.

4. OBJECTIVE main hypothesis underlying our project is that the visual systems of ocean creatures have evolved to adapt to the natural statistics of the aquatic scenes.

5. INTRODUCTION Since oceans, rivers and lakes contain abundant resources, underwater imaging has become an important researching filed and received much attention recently. While due to the absorption and scattering when light is traveling in water, there are three major problems of underwater imaging: colour distortion, under-exposure and fuzz. This degradation is mainly caused by the physical properties of the medium.

6. Retinex and noise aware shadow-up BLOCK DIAGRAM FIG: BLOCK DIAGRAM OF RETINEX AND NOISE AWARE SHADOW -UP

7. Decomposition based on Retinal theory Retinex theory: Original image (s)= illumination (L)* reflectance (R) Need to estimate these ‘L’ and ‘R’. In general in camera representations halo antifacts occur unnaturally in the boundary of regions with large gradient values.

8. WEIGHTED VARIATIONAL MODEL Weighted variational model (WVM) was proposed for simultaneous reflectance and illumination estimation. Input only ‘V’ we are taking here, R=has almost no noise; L=includes noise

9. WEIGHTED VARIATIONAL MODEL S=R.L; R [0 1] L[0 INFINITE] Logarithmic approach to find R, L Log s= log R+ log L illumination is its spatial smoothness S<L ( Bcz R=point val  point val *L lesser than L) If R=1 S=L S≤L By keeping l as constant find R check output.

10. adaptive gamma correction with weighting distribution (AGCWD) Used to enhance the image value of gamma is find out automatically with the help of weighted distribution function. Data Transformation using AGCWD T (l) = cdf (l) lmax . n=0.8

11. PROBABILITY DISTRIBUTION FUNCTION pdf= Probability distribution function (PDF) of luminance is the probability that a brightness chosen from the region is less than or equal to a given brightness value ‘a’. Pdf (l) = n1 / (M*N) n1=number of pixels that have intensity l MN =total number of pixel.

12. Shadow-up function to avoid the loss of details in bright areas due to over enhancement. Here in this process, we can take only enhanced pixel if original pixel value is greater than Ith ; else we can take direct original pixel I(x, y)∈[0,255]=intensity of illumination layer T(I( x,y ))= AGCWD output. We are taking enhanced AGCWD op iff I( x,y )< Ith I th =255-(sum(all pixel intensities)/No. of pixels)

13. YUV (luma, chroma values) Output image = Enhanced illumination image *Reflectante value and V’=I′(x, y)·R(x, y). Convert into YUV Y′UV model defines a colour space in terms of one luma component (Y′) and two chrominance components, called U (blue projection) and V (red projection) respectively

14. Block-matching and 3Dfiltering (BM3D) Only ‘y’ channel is applied to BM3D to remove noise. Block-matching: used to group similar property regions. Image fragments are grouped together based on similarity. Fragments do however have the same size. A fragment is grouped if its dissimilarity with a reference fragment falls below a specified threshold. This grouping technique is called block-matching. BM3D on the other hand may group macroblocks within a single frame. All image fragments in a group are then stacked to form 3D cylinder-like shapes.

15. ADVANTAGES 1. More accurate with high reliability 2. Fast processing 3. support any type of format

16. drawbacks We can not get clear images in fog. Disturbances due to dust . Video surveillance systems, obstacle detection systems, outdoor object recognition systems and intelligent transportation systems can not be operated.

17. APPLICATIONS 1. Automobiles 2. Navy, aviation's 3. Space applications 4. Submarines

18. RESULTS HAZE INPUT IMAGE

19. RESULTS Depth Maps

20. RESULTS Transmission Maps

21. RESULTS OUTPUT DEHAZE IMAGE

22. CONCLUSION Finally, we demonstrate the values of modelling the visual mechanisms of underwater creatures for the challenging underwater image processing task. We introduced a non uniform colour correction algorithm, which could well handle the non uniform colour cast in underwater images compared to the existing methods. We imitated the adaptive retinal mechanisms to control the model parameters of each low level filter according to the global contrast of a given image. We exploited the colour-opponent mechanisms to flexibly adjust the colour appearance of underwater images during image enhancement.

23. FUTURE SCOPE To enhance the visibility of image caused by atmosphere suspended particles like dust, haze and fog which causes failure in image processing such as video surveillance systems, obstacle detection systems, outdoor object recognition systems and intelligent transportation systems. And visibility restoration techniques should be developed to run under various weather conditions.

24. references [1] W. Jacobs, V. Nietosvaara, A. Bott, J. Bendix, J. Cermak, M. Silas, and I. Gultepe , “Short range forecasting methods of fog visibility and low clouds,” Earth System Science and Environmental Management Final Rep. on COST-722 Action, 2007. [2] D. Zhang, E. O’Connor, T. Sullivan, K. McGuinness, F. Regan, and N. E. O’Connor, “Smart multi-modal marine monitoring via visual analysis and data fusion,” in Proceedings of the 2nd ACM international workshop on Multimedia analysis for ecological data, pp. 29–34, ACM, 2013. [3] D. Zhang, T. Sullivan, C. C. Briciu Burghina , K. Murphy, K. McGuinness, N. E. O’Connor, A. F. Smeaton, and F. Regan, “Detection and classification of anomalous events in water quality datasets within a smart city-smart bay project,” International Journal on Advances in Intelligent Systems, vol. 7, no. 1&2, pp. 167–178, 2014. [4] S. A. Chatzichristofis and Y. S. Boutalis , “ Cedd : color and edge directivity descriptor: a compact descriptor for image indexing and retrieval,” in Computer Vision Systems, pp. 312–322, Springer, 2008. [5] S. Park, D. Park, and C. Won, “Core experiments on mpeg-7 edge histogram descriptor,” MPEG document M, vol. 5984, p. 2000, 2000.