Signal detection performance by a human observer is often the gold standard in assessing the quality of images and hence the best imaging system. Several extensive mathematical models have been proposed to imitate human observer performance. This is critical in areas like imaging (imitating a radiologists), defense (finding signals in dense or turbulent atmospheres), security (finding explosives from x-ray images). We show methods based on simple well known texture metrics that can correlate well with human observer detection performance. Our claim is that these methods when weighed suitably can be an excellent replacement for complicated and time consuming human observer studies aimed at designing and optimizing imaging systems, software such as filters and algorithms used to smooth or improve images or any such operations that may be applied to the images. Our results for low contrast signal detection in breast images show excellent proof of concept.