The study and analysis of facial expressionhaspaying attention tremendouslyas human being is veryconscious inthedomainof computer vision because ittakesaprincipalrole in thearea ofman-machinecommunication.Almost all types offace expressions arebasically spawnedbytaking a new hueof facialmuscles, whichgradesin temporallymalformedfacial features such asforehead,eye-lids, eye-brows, nose,lips, and skincolour/texture, etc.During the evaluation procedure of any facial expression emphasiseshould have to be given on the following parameters :( a)the partsof the face that willplay prime role infacial action,(b)the intensity of facial actions, and(c)thestochastic parametersof facial actions.In ourpaper wehave proposedareal-time,precise-exact, and robustfacial expressionrecognitionmethodologydependingon genetic algorithmalong withWeighted Vector DirectionalFilters (WVDF) as it is provedthat it is the most powerful unbiased optimization techniques for sampling a huge solution space. The GAcan be optimized by using WVDF filter'scoefficients. By using the Stochastic Selection of parents and thegeneration of optimal offspring, the optimal WVDF filter can be obtained.Wehave been capable toconstruct aleaf-matrix toour requiredintendedexpression.Ultimately, we haveevaluatedourproposedmethodology by using Weighted Vector DirectionalFilters (WVDF) along with genetic algorithm with theshutter function of Canon Camera. We haveaccomplished our job by achievingbetter performance thanSony on slight function.
展开▼
