FPGA Implementation of Medical Image Fusion Based on DWT using Spartan 3EDK

FPGA Implementation of Medical Image Fusion Based on DWT

Overall rating

Now-a-days, almost all areas of medical diagnosis are impacted by the digital image processing. When an image is processed for visual interpretation, the human eye is the judge of how well a particular method works. Clinical application demanding Radiotherapy plan, for instance, often benefits from the complementary information in images of different modalities. For medical diagnosis, Computed Tomography (CT) provides the best information on denser tissue with less distortion. Magnetic Resonance Image (MRI) provides better information on soft tissue with more distortion. With more available multimodality medical images in clinical applications, the idea of combining images from different modalities become very important and medical image fusion has emerged as a new promising research field. The experiments show that the method could extract useful information from source images to fused images so that clear images are obtained. In this paper, a hardware implementation of a image fusion system is proposed. Here MATLAB is used to convert images into pixel-format files and to observe simulation results. To implement this paper XPS & VB are needed. In XPS, first select hardware & software components then by adding source and header files & converting into bit streams and download into FPGA, to obtain fused image. The input image can also be recovered by combining of fused image and the other input image.

Proposed Method

The Discrete Wavelet Transform [7] was developed to apply the wavelet transform to the digital world. Filter banks are used to approximate the behaviour of the continuous wavelet transform. The signal is decomposed with a high-pass filter and a low-pass filter. The coefficients of these filters are computed using mathematical analysis and available in subsections. A. Wavelet Transform for Fusing Images: In this subsection, the detailed fusion steps based on wavelet transform can be summarized as shown in below Fig.1.

  • First register the input images (I1 and I2), which are going to be fused and corresponding pixels are aligned.
  • The Registered input images are decomposed into wavelet transformed images respectively, based on haar wavelet transformation (W). The transformed images with K-level decomposition will include one low-frequency portion (LL band) and 3K high-frequency portions (LH bands, HL bands and HH bands).
  • The Transform coefficients of different portions are performed with a certain fusion rule.
  • Then by applying performing an inverse wavelet transform (W-1) based on the combined transform coefficients, the fused image (I) is constructed.

Wavelet Decomposition of Images

Filter banks decompose the signal into high- and low-frequency components. The low-frequency component usually contains most of the frequency of the signal. This is called the approximation. The high-frequency component contains the details of the signal. Wavelet decomposition can be implemented using a two-channel filter bank. In wavelet decomposing of an image, the decomposition is done row by row and then column by column. For instance, here is the procedure for an N x M image. Then filter each row and down-sample to obtain two N x (M/2) images, they are L and H images. The formulas used to do find H and L is given below. H= odd-even = F1; L= even + round (F1/2); Then filter each column and subsample the filter output to obtain four (N/2) x (M/2) images, by using the formulas shown in below. LH= odd – even = F2; LL= even + round (F2/2); HH= odd – even = F3; HL= even + round (F3/2); The four sub-images obtained as seen in Fig. 1, the one obtained by low-pass filtering the rows and columns is referred to as the LL image. Similarly LH, HL and HL are also formed. Each of the sub images obtained in this fashion can then be filtered and subsample to obtain four more sub images. This process can be continued until the desired sub band structure is obtained. Here F1,F2 and F3 are taken as references.

  • In the structure of pyramid decomposition, only the LL sub image is decomposed after each decomposition into four more sub images.
  • In the structure of wavelet decomposition, each sub image (LL, LH, HL and HH) is decomposed after each decomposition.
  • In the structure of space, after the first level of decomposition, each sub image is decomposed into smaller sub images, and then only the LL sub image is decomposed.


Join the World's Largest Technical Community

we respect your privacy.