Font Size: a A A

The Research On Multi-focus Image Fusion

Posted on:2009-10-16Degree:MasterType:Thesis
Country:ChinaCandidate:X FengFull Text:PDF
GTID:2178360245953666Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Image fusion is the process of combining two or more images captured from different kinds of sensors into an image. And this process can obtain resultant images with higher confidence, less blur, more intelligibility and is suitable for human vision and computer processing such as detection, classification, recognition and comprehension.At present, a wide variety of data acquisition devices are available. There are sensors which can not generate images of all objects at various distances (from the sensor) with equal clarity (e.g. camera with finite depth of field, light optical microscope, etc.). Under this circumstance, several images of a scene are captured, with focus on different parts of it. The acquired images are complementary in many ways and a single one of them is not sufficient in terms of their respective information content. However, viewing a series of such images separately and individually is not very useful and convenient. The advantages of multi-focus data can be fully exploited by integrating the sharply focused regions seen in the different images.In this paper, we proposed two multi-focus image fusion approaches based on the wavelet analysis, genetic algorithm and morphological operation.The first one divides the source images into square blocks, introduce the eigenvalues to assess the blocks quality, say clear or blurry. Then select the clear blocks to combine the resultant image. The genetic algorithm based on integer is employed to optimize the best sizes of the blocks.The second algorithm firstly applies one-scale wavelet decomposition to the source images, conducts clear region detection on the detail images, finally combine the clear region detected into the resultant image. GA is also used to select the concerned parameter in the algorithm.These algorithms are both region based image fusion algorithms, which lies between pixel level image fusion and feature level image fusion. Experimental results demonstrate that the proposed schemes outperforms Haar wavelet approach and morphological wavelet approach, both in visual effect and objective evaluation criteria, particularly when there is movement in the objects or mis-registration of the source images.
Keywords/Search Tags:Image Fusion, Multi-focus Image, Genetic Algorithm, Wavelet Analysis, Quality Assessment
PDF Full Text Request
Related items