The software quality can be enhanced with the awareness and compassionate about the software faults. We acknowledge the impact of threshold of the object-oriented metrics on fault-proneness. The prediction of fault-prone classes in early stage of the life-cycle assures you to allocate the resources effectively. In this paper, we proposed the logistic regression based statistical method and metric threshold to reduce the false alarm for projects that fall outside the risk range. We presented the threshold effects on public datasets collected from the NASA repository and validated the use of threshold on ivy and jedit datasets. The results concluded that proposed methodology achieves the speculative results with projects having similar characteristic.

Conventional visual secret sharing schemes generate noise-like random pixels on shares to hide secret images. It suffers a management problem, because of which dealers cannot visually identify each share. This problem is solved by the Extended Visual Cryptography scheme (EVCS). However, the previous approaches involving the EVCS for general access structures suffer from a low contrast problem. This paper proposes a new (k,n)-threshold image sharing scheme using extended visual cryptography scheme for color images based on bit plane encoding that encrypts a  color image in such a way that results of encryption is in the form of shares. Shares do not reflect any information directly, information is scrambled instead. The traditional binary EVCS is used to get the sharing images at every bit level of each principle component of a color image. This scheme provides a more efficient way to hide natural images in different shares. Furthermore, the size of the hidden secret can be recovered by inspecting the blocks in the shares. This new scheme for color images gives the ideal contrast in the recovered image.