The present invention is related to agricultural land monitoring using image processing. More particularly, the invention is related to elimination of speckle 5 noise present in agricultural SAR images. The introduction of speckle noise mainly occur in image aquasition phase. Radar imagery are especially used during monsoon season. Satellite images are obtained from synthetic aperture radar, that is fixed on the aircraft that captures the high resolution images of the broad areas of the earth surface. SAR images are formed by the consistent interaction of the 10 emitted microwave radiation with target areas. This consistent interaction originates arbitrary constructive and destructive nosiness resulting into multiplicative kind of noise known as speckle noise all over the image. Most of the restoration models are additive in nature. So, nature transformation is needed from multiplicative to additive by which the available effective noise restoration 15 models can be used effectively. Apart from agricultural land monitoring, it also helps in wave forecasting, marine climatology, regional ice monitoring, ship detection in the coastal regions and field of tropical forest monitoring.
BACKGROUND & PRIOR ART 20
There are many despeckling techniques and filters available in spatial as well as frequency domain to handle speckled SAR images by preserving the major information in the image like edge, texture, homogeneous and non-homogeneous regions of the image. Some standard and conventional method for despeckling purpose in the field of satellite imagery are, frost filter, kuan filter, kuwahara 25 filter, lee filter, mean filter and median filter. Other developed schemes under homomorphic and non-homomorphic filtering are effective and adaptive. It is observed that bayesian approaches in transform domain shows better results than bayesian approaches in spatial domain. But there some non-bayesian approaches that gives as better results as bayesian approaches in transform domain. Order 30
3
Statistics and Morphological Filters are helpful in preserving edges, retaining texture and smoothing the noisy background, but thicker objects left unprocessed. This method is not specifically built for speckle noise model but still provides fair results in some cases. Anisotropic diffusion is a popular technique that retains significant parts mainly edges, lines and other fine details. Speckle reduction 5 using anisotropic diffusion exploits sudden occurring coefficient of variation. It shows better results than conventional methods in the terms of variance minimization, mean preservation and edge localization. In despeckling based on compressed sensing it is a known that in order to obtain good quality image, multiple degraded images can be merged. Keeping this in mind, compressed 10 sensing is employed to get multiple SAR images from a single SAR image. Other non-bayesian approaches like bilateral filtering, sigma filter and non-local filtering which also gives satisfactory results in terms of visual appearance and edge preservation. Homomorphic and non-homomorphic filtering are the two bayesian methods in transform domain that provides best schemes in wavelet 15 domain. Homomorphic filtering is used to remove the multiplicative noise using log and exponential operations, while non-homomorphic filtering is less frequently seen in literature due to its complexity of directly handling the multiplicative noise. Homomorphic filtering is in much use since last two decades as after transforming multiplicative noise to additive, other additive noise models 20 can be used easily and effectively to handle the situation, it is easy method to understand, while non-homomorphic filtering methods directly works upon multiplicative noise. This method is comparatively difficult to work upon and is less effective too. Basically homomorphic filtering is used for improving non-homogeneous illumination in images. Classical hard and soft thresholding 25 methods were implemented in. The undecimated wavelet transform and the MAP standard have been implemented in the issue of SAR image despeckling.
OBJECTIVE OF THE INVENTION
The principal objective of the present invention are:
 Removal of granular structured speckle noise. 30
4
 Preservation of edges, texture, homogeneous and non-homogeneous regions of agriculture land.
 National agricultural drought assessment and monitoring.
 Country-wide agricultural land-use mapping.
 Crop production forecasting. 5
 Crop damage-assessment.
 Crop planning and diversification.
BRIEF DESCRIPTION OF DRAWINGS
The present invention eliminates the granular structured speckle noise which occurred during image acquisition. Before performing any type of image 10 processing operation like, segmentation, morphology, recognition etc, image enhancement and denoising is the first preprocessing step to be performed . The present invention uses homomorphic filtering (2, 38) in order to handle multiplicative nature of speckle noise. The multiresolution decomposition is performed using db2 based DWT upto level 3 (3-8). This decomposes the 15 speckeld SAR image into two components, approximate (3, 5, 7) and detailed component (4, 6, 8). The approximate component (7) is subjected to NLM filter (9) and NLM filter (24) is also applied during inverse DWT. All the three detailed components (4, 6, 8) are subjected to method noise wavelet thresholding (31, 23, 16). The despeckled image (32) is achieved after IDWT. With the intention to 20 enhance the results, method noise wavelet thresholding (37) is again applied on despeckled image (32). This whole procedure finally provides the despeckled output image (39).
DETAIL DESCRIPTION
In order to achieve an good visual quality image, a method noise wavelet 25 thresholding (31, 23, 16, 37) and homomorphic filtering (2, 38) based despeckling framework using NLM filter (9, 24) is proposed which generates no artifacts and a brilliant visual appearance of an output image. Speckle noise is multiplicative in nature. Therefore the effect of distortion produced by this noise is high as
5
comparison to additive noises. The homomorphic filtering (2, 38) is performed to handle the multiplicative nature of the speckle noise. The homomorphic filtering (2, 38) is used to correct non-uniform illumination and to enhance contrasts in the image. It’s a frequency filtering, preferred to other techniques because it corrects non-uniform lightening and sharpens the edges at the same time. 5
It adopts the reflectance and illumination models. Since illumination and reflectance combine multiplicatively, the components are made additive by taking the logarithm (2) of the image intensity, so that these multiplicative components of the image cane be separated linearly in the frequency domain. Similarly log transform (2) is applied over the speckled SAR image to change its nature to 10 additive. This homomorphic (2, 38) phase helps to use the best pre-developed additive restoration models. It minimizes the time and complexity of the work while enhances the simplicity and quality of the work. Wavelets are preferred mathematical function for despeckling purpose. DWT provides the transformation of the SAR image from the spatial to the frequency domain. MATLAB wavelet 15 toolbox offers a function, dwt2 for 2D-DWT which is used to analyse the high-frequency component in the image. Daubechies wavelet (db2) is used as it easily solves the self-similarity properties of a signal or fractal problems. Decomposition is tested and performed up to 3 levels for visual quality requirement analysis. DWT decomposes the image into approximate part (3, 5, 7) and detailed part 20 [horizontal, vertical and diagonal] (4, 6, 8). The approximate part (7) is subjected to NLM filter (9).
25
Sign: -
Name: - SHREE, RAJ
Address: - B-1/1-4, DEPARTMENT OF INFORMATION TECHNOLOGY, BABASAHEB BHIMRAO AMBEDKAR UNIVERSITY, VIDYA VIHAR, LUCKNOW (226025), UP, INDIA
Sign: -
Name: - SINGH, PRABHISHEK
Address: - G 642, AWAS VIKAS 1, KALYANPUR, KANPUR (208017),
UP, INDIA
Sign: -
Name: - PANDEY, RAVI PRAKASH
Address: - 5/5/21A, NEAR UDAYA PUBLIC SCHOOL, ITI CHAURAHA, PANCHKOSHI PARIKRAMA MARG, FAIZABAD, UP, INDIA-224001
6
The detailed part (4, 6, 8) is thresholded using bayesian shrinkage rule using soft thresholding (25, 17, 10) followed by method noise wavelet thresholding (31, 23, 16). The concept of method noise is new and introduced in last few years. Method noise is highly adaptive in any application of image restoration. NLM filter (24) is also applied during IDWT in FIG 1. This whole procedure gives S’ (32) as 5 despeckled image.
In any denoising method, complete denoising is not possible, some sort of high frequency areas retains in the output image. In order to handle this situation, method noise thresholding is again applied over S’ (32). Finally, a granular structured speckle free output SAR image, S’’’ (39) is obtained.

CLAIMS
The despeckling SAR image method comprising:
1. provides a unique and highly adaptive method of despeckling the speckled SAR images in two different scenarios. First, when speckle free original SAR image is available as reference image and speckle noise is added to it 5 to test the algorithm and second case is when reference image is real speckled SAR image. The hybrid method is unique and highly capable of despeckling the SAR image in both the cases efficiently.
2. uses the concept of “method noise thresholding” in despeckling field using db2 based 2D-DWT. 10
3. provides a unique and intelligent way of filtering the detailed part of the speckled SAR image which holds the maximum information by applying wavelet thresholding followed by method noise thresholding on the detailed part of the image.
4. provides a unique and hybrid combination of non-local mean filter, 15 wavelet thresholding and method noise thresholding for smoothing the agricultural SAR image using db2 based 2D-DWT for significant speckle reduction.