The machines and robots interact with thehuman with the help of embedded system. for example, with the help of handgesture machine and human interaction. This interaction can be possible withthe implementation of the embeddedsystem.
With the advancement in technology,the machine vision is used to identify the various human gestures, in somesystem wheeled robots can be used toidentify the human gesture. Some embedded system would not support the usage ofKinect due to its volume and weights. The computation time related to thisembedded platform is high. Therefore, it is necessary to reduce the computationtime of these embedded systems. This can be reduced by categorizing the skinareas and non- skin areas of the hand. The significance of this study is toquantitatively evaluate the proposed hypothesis to fill the identified researchgaps related to proposed methodology.
The gestureis defined as the way that can be used tounderstand the significant changes related to hand moments that can be possiblewith the help of embedded systems. The embedded systems. Embedded system is computers that can be used for real timecomputing. the embedded systems are integrated with other system and thatinclude mechanical and hardware parts. As compared to general computers theembedded system consumes less power, consumes less cost, small in size andrugged operating system that can be used to perform real time computations.
Thehuman gestures can be recognized with the help of machine learning algorithmsand classification of the images can be done according to the various featuresof the image. The classification of the image can be done according to the skin area ofand non-skin area of the hand. The areseveral types of embedded system that can be useful in our daily life but inour approach, we have used these systemsfor identifying the human gestures. Statement of problem and sub problemsThis research is conducted to reduce thetime of the various computational statistics that are involved whilerecognizing the hand gesture.
Some embedded system does not support the usageof Kinect due to its volume and weights. The computation time related to thisembedded platform is high. Therefore, it is necessary to reduce the computationtime of these embedded systems. This can be reduced by categorizing the skinareas and non- skin areas of the hand.The real time recognition can be done by using Gaussian mixture model. thistypes of intelligent robots can be used for security and privacy purpose.
Following are the sub-problems of the proposedalgorithm:Sub-problem1:The accuracy rate of the proposedalgorithm to analyze the input image: The problem related to accuracy anddetection rate of images has been analyzed with the help of quantitativeanalysis. The detection of images can be done in a reduced number of computation related to conventional approaches. Limitations of studyFollowingare the limitations of conducted research: · The result evaluation andanalysis of proposed methodology is out of the scopeof this study. · Large scale dataset isnot included in this study. Gaussian mixture model algorithm is evaluated withthe help of a limited number of images.· The average rate ofrecognition related to adopted algorithm or approach is 75%. It can be improved.
The images recognized with the help of proposed recognition are less. It can beimproved by using a number of machinelearning algorithms. · The data input to theproposed approach is not variant. Variant data can be used to evaluate theoriginality of algorithm. the number of resources from the data is collected islimited.
More will be the data more will be originality of conducted research. HypothesisThe hypothesis related to the reduction of computational time can beevaluated with the help of proposed analysis. it is assumed that system wouldbe capable of detecting the various hand gesture in an effective manager ascompared to conventional approaches. following is the hypothesis of proposedapproach:H1: The computational time can be reducedwith the implementation of proposed algorithmH2: The computational time cannot be reduced withthe implementation of proposed algorithmH3: The proposed system can detect thehand gesture with the help of Gaussian mixture model