In this study, blind code extraction of Direct Sequence Code Division Multiple Access (DS-CDMA) signals is considered, based on Independent Component Analysis (ICA) method. In order to distinguish between correct and incorrect extracted codes, to estimate the number of active users and also to determine the quality of detection along with the ICA based blind detection procedure, some propositions are defined. These propositions are used to improve the performance of the ICA blind detection based method. Then, in order to analyze the proposed criteria, Principle Component Analysis (PCA) and Gaussian Mixture Model (GMM) are employed. Experimental results illustrate that the achieved performance of the defined criteria.