سال انتشار: ۱۳۸۹
محل انتشار: ششمین کنفرانس ماشین بینایی و پردازش تصویر
تعداد صفحات: ۴
sepideh almasi – control and intelligent processing center of excellence school of electrical and computer engineering university of tehran
hamid soltanian-zadeh – image analysis laboratory radiology department
A methodical definition of the probability of error for each of compressive sensing methods a performance evaluating tool that has not been defined before , is the theme of this work. To this end an efficient information theoretic method is proposed in order to acheive a reliable measure with making a replication with the communication channels, the probability of error is set as a function of the mutual information MI with the compressive sensing process . other methods of performance assessment including different kinds of error such as mean square error MSE are in use prevalently however fail in some cases because of the lack in inclusion of the coherency between structure of the result and that of the original signal.