Abstract:

The study is based on the use of generalized Multiple Attractor Cellular Automata (GMACA). GMACA is a type of cell automates applied to more than one cell. The application of cellular automates to more than one cell is performed using the rule vector. In literature research, the task of recognizing human movements in video images; what it will contain, the importance, applications, etc. The topics are investigated. The recognition of human movements is carried out in five stages; it is necessary to identify and track the human object, to extract the characteristics of the human object and to use these characteristics to extract the characteristics of the movement, to recognize the activities consisting of simple movements. In the developed method of movement recognition, the images are first converted into a grey color space. The GMACA rule vector will then be created to be used for movement detection. The movement is detected using GMACA. The HOG character vector is removed from the movement detection images and the HOG character vectors obtained are labeled according to the movement they belong to. This is how the data set is created. The data set created is divided into training and test data sets by cross-right method. The recognition of human movements is performed by the SVM method. Experimental results are shown with the matrix of the mixture. The classification performance of the recognition method developed using the mixture matrix has been shown. The motion recognition application made with the movement detection images obtained by GMACA has given the same good results as the existing background drawing work. The results show that GMACA can be used in movement detection and movement recognition studies. The weakness of GMACA is that it can be applied on binary patterns. The movement detection method we have developed can be applied after the pixels in the decimal base are converted into a binary base.

Keywords: