DESIGN OF FEATURE EXTRACTION TECHNIQUE FOR UNDERWATER IMAGE FUSION

Abstract

Submerged pictures typically need differentiate and experience the ill effects of shading mutilation because of light bar dissipating and lessening. Light dispersing is because of the presence of sispended particles in water in type of both natural and inorganic material which mirrors and diverts the light in a capricious way before it arrives at the sensor and results in a picture which is low conversely. Water as a medium promptly retains light, and in addition various frequencies of light as ingests at various rates. Moreover, the more extended frequency is assimilated first and it brings about progress upon the entrenched submerged picture combination technique. our submerged picture upgrade framework can comprehensively be separated in three significant segments; parting of the first picture into illuminance and reflectance cuts, applying a straight piecewise shading amendment calculation on the reflectance cut and agreement improving methods on the illuminance part. At last, to remake the yield picture we apply the multi stage combination strategy which depends on SIFT and HOG. Exploratory outcomes ahow an engancement for the effectively accessible combination based method. Additionally we furnish an examination with two different strategies. The performance of the proposed method is compared with an existing method in terms of Mean Square Error and Peak Signal to Noise Ratio, which indicates that the proposed method is better than the existing method.