الفهرس 
Infrared Image Enhancement TechniquesOnly 14 pages are availabe for public view
 |  | from | 100 |  |  |
| | | from | 100 | | |
Abstract
Recently, Infrared (IR) Images attract the attention of most researchers due to their importance in many areas like medical, military, and night vision. IR images suffer from some problems like low Peak Signal to Noise Ratio (PSNR), low Resolution, low contrast, and difficulty to detect objects. Also, IR images have large black areas and small details due to the absence of light needed for imaging. These problems limit the overall performance of IR images. So, it is necessary to enhance the IR images. This thesis presents two efficient approaches for enhancing the IR images.
The first proposed approach is ―Quality Enhancement of Infrared Images Based on Dynamic Fuzzy Histogram Equalization and High Pass Adaptation in Discrete Wavelet Transform (DWT)‖. This approach presents an efficient method for enhancing the contrast, appearance, and quality of IR images. The proposed approach works by applying DWT to the IR image to obtain wavelet coefficients. The low frequency band LL is enhanced by employing DFHE. HPA is applied to the high frequency band HH and the middle frequency bands LH and HL. Experimental results demonstrate that the proposed approach is very efficient in enhancing contrast and sharpening the edges of the low contrast IR images. The second proposed approach is ―Enhancement of Infrared images using homomorphic filtering in Fast Discrete Curvelet Transform‖. This approach contributes also for enhancing the quality of IR images by mixing the advantages of FDCT for representing curves and clarifying features on it with homomorphic filtering which is an efficient approach for enhancing the appearance and contrast of IR images.
Also, the second proposed approach investigates the influence of FDCT coefficients on the IR images enhancement. When using all FDCT coefficients, the image quality is enhanced greatly. But it takes long time to implement this
approach due to the nature of the FDCT transform because of complex calculation. Using only two coefficients of FDCT enhance the IR images and takes short time. And this can be used in applications like military, medical. However an application such as temperature forecasting time is not effect greatly. So, the proposed approach with all FDCT coefficients can be used. Finally, a comparative study is employed to compare the two proposed approaches with other traditional enhancement approaches. Results demonstrate that the first proposed approach is the best one for enhancing the IR image since it enhances contrast, sharpening edges and image’s features, average gradient, standard deviation, and noise reduction. The second proposed approach is better for enhancing PSNR, edges, curves, and clarifying features in the IR images.
Different IR databases are added to test the efficiency of the two proposed approaches.
Finally, noise is added to the IR images to show the effect of noise on the performance of the two proposed approaches. The first proposed approach handle noise in efficient way. The second proposed approach needs a pre-processing step before enhancing IR image to give better results.