روشي جهت تعيين پارامتر بهين هي فليترهاي نرم به منظور تلفيق تصاوير در فضاي فركانس

A method for determining the optimum parameter of the soft filters to image fusion in the frequency domain

ﺗﻠﻔﻴﻖ ﺗﺼﻮﻳﺮ راﻫﻜﺎري راﻳﺞ ﺑﻪ ﻣﻨﻈﻮر اﻳﺠﺎد ﻫﻢ اﻓﺰاﻳﻲ در ﻣﺤﺘﻮاي ﺗﺼﺎوﻳﺮ ﭼﻨﺪﻃﻴﻔﻲ و ﭘﺎﻧﻜﺮوﻣﺎﺗﻴﻚ ﻣﺤﺴﻮب ﻣﻲ ﺷﻮد. ﺗﺎﻛﻨﻮن راﻫﻜﺎرﻫﺎي ﻣﺨﺘﻠﻔﻲ ﺑﻪ ﻣﻨﻈﻮر اﻧﺠﺎم ﻓﺮاﻳﻨﺪ ﺗﻠﻔﻴﻖ ﺗﻮﺳﻌﻪ ﻳﺎﻓﺘﻪ اﻧﺪ ﻛﻪ ﺑﻪﻛﺎرﮔﻴﺮي ﻓﻀﺎي ﻓﺮﻛﺎﻧﺲ ﻳﻜﻲ از اﻳﻦ راﻫﻜﺎرﻫﺎ ﺑﺸﻤﺎر ﻣﻲ رود. ﺗﻠﻔﻴﻖ ﺗﺼﺎوﻳﺮ در ﻓﻀﺎي ﻓﺮﻛﺎﻧﺲ ﺑﻪ ﻛﻤﻚ ﻓﻴﻠﺘﺮﻫﺎي ﺑﺎﻻﮔﺬر و ﭘﺎﻳﻴﻦ ﮔﺬر اﻧﺠﺎم ﺷﺪه ﻛﻪ ﺗﻌﻴﻴﻦ اﺑﻌﺎد ﺑﻬﻴﻨﻪ ي ﻓﻴﻠﺘﺮﻫﺎ ﻣﻬﻢﺗﺮﻳﻦ ﭼﺎﻟﺶ آن ﻣﺤﺴﻮب ﻣﻲ ﺷﻮد. در اﻳﻦ ﻣﻘﺎﻟﻪ ﺑﻪ ﻣﻨﻈﻮر ﺑﻬﺮهﻣﻨﺪي ﺗﻮأﻣﺎن از ﻣﺤﺘﻮاي ﻃﻴﻔﻲ و ﺟﺰﺋﻴﺎت ﻣﻜﺎﻧﻲ ﺗﺼﺎوﻳﺮ، ﺷﺎﺧﺼﻲ وزن دار ﻣﺘﺸﻜﻞ از دو ﺷﺎﺧﺺ رﻧﮓ ﭘﺬﻳﺮي و ﺟﺰﺋﻴﺎت ﻣﻜﺎﻧﻲ ﭘﻴﺸﻨﻬﺎد ﺷﺪه اﺳﺖ. ﺷﺎﺧﺺ ﻣﺬﻛﻮر ﺗﻮاﻧﺴﺘﻪ ﺑﺎ اﺗﻜﺎ ﺑﻪ ﻣﻌﻴﺎرﻫﺎي ﻫﻨﺪﺳﻲ و آﻣﺎري ﻓﻀﺎي وﻳﮋﮔﻲ در ﻛﻨﺎر ﺑﻪﻛﺎرﮔﻴﺮي ﻓﻴﻠﺘﺮﻫﺎي ﻧﺮم، ﻧﺘﺎﻳﺞ ﻣﻄﻠﻮﺑﻲ را در ﺷﻨﺎﺳﺎﻳﻲ اﺑﻌﺎد ﺑﻬﻴﻨﻪ ي ﻓﻴﻠﺘﺮﮔﺬاري ﺗﺄﻣﻴﻦ ﺳﺎزد. در ﺳﺎزوﻛﺎر ﭘﻴﺸﻨﻬﺎدي، وزن ﺷﺎﺧﺺ ﻫﺎي رﻧﮓ ﭘﺬﻳﺮي و ﺟﺰﺋﻴﺎت ﻣﻜﺎﻧﻲ ﺑﺎ در ﻧﻈﺮ ﮔﺮﻓﺘﻦ ﺗﻤﺎﻳﺰ در ﻣﺤﺘﻮاي ﺗﺼﺎوﻳﺮ، ﺑﺮاي ﻫﺮ ﺗﺼﻮﻳﺮ ﺑﻪ ﻃﻮر ﻣﺴﺘﻘﻞ ﺑﺮآورد ﻣﻲ ﮔﺮدد. ﻣﻘﺎﻳﺴﻪ ي ﺑﻬﺘﺮﻳﻦ ﻧﺘﺎﻳﺞ ﻛﺴﺐ ﺷﺪه از روش ﭘﻴﺸﻨﻬﺎدي ﺑﺎ روش ﻫﺎي راﻳﺞ ﺗﻠﻔﻴﻖ ﺣﺎﻛﻲ از ﺑﻬﺒﻮد ﺑﻪ ﻃﻮر ﻣﺘﻮﺳﻂ 58 درﺻﺪي در دﻗﺖ )RMSE ( ﻓﺮاﻳﻨﺪ ﺗﻠﻔﻴﻖ اﺳﺖ.

Image fusion is known as a synergetic process for merging multispectral and panchromatic images contents. So far, various methods have been developed in which the usage of the frequency domain is one of them. The frequency-based image fusion techniques are performed using high and low pass filters. So, the determination of the sizes of these filters would be a challenge. In this paper, a weighted index is proposed to determine the sizes and shapes of the low and high filters in fusion of the panchromatic and multispectral images. In the proposed method, the weights of the spectral and spatial indicators are independently estimated for each image. So, the effects of the differentiation of the image contents and different range of the indicators are properly adjusted to reach the optimum filtering. The comparison of the best results obtained from the proposed method with the other well-known fusion methods, in the used datasets, was indicated an average improvement of 58% in RMSEs.