Computer software has progressively turned out to be an essential component in modern technologies. Testability is the probability whether tests will detect a fault, given that a fault in the program exists. How efficiently the faults will be uncovered depends upon the testability of the software. Various researchers have proposed qualitative and quantitative techniques to improve and measure the testability of software. Furthermore, it is well known fact that the progress of software testing is influenced by various uncertainty factors like effort expenditure, skill of test personal, testing tool, defect density and irregular state of software fault-report phenomena on the bug tracking system. Hence, there is an irregular fluctuation in fault detection/removal rate during testing phase. In this paper, distribution based software reliability growth models have been developed by applying Itô type Stochastic Differential Equations in order to incorporate (i) the irregular fluctuation in the fault detection process due to various uncertainty factor during testing phase; (ii) two types of imperfect debugging and (iii) change-point concept. The proposed stochastic differential equation based models have been validated using real data sets. Various comparison criteria results for goodness of fit have also been presented in the paper.