The paper presents a changed method based on likelihood ratio test in combination with Bayes' theory developed for foliage penetration (FOPEN) ultra-wideband (UWB) lowfrequency SAR systems. The method is tested on new experimental data collected by the CARABAS II/LORA-VHF system in 2021 in dense forests. The detectability is 87.5% of the nine targets deployed and the false alarm rate is 0.44 false per square kilometer. The probabilities of the detected targets are between 95% and 100%. Traditionally, for SAR change detection, we can calculate a likelihood ratio, giving the range of values of a ratio of two probabilities from 0 to infinity. In this study, we transform the likelihood ratio test to a conditional probability based on Bayes' theorem. With such a transform, the change detection results will be represented by the detected changes, and the detected changes will be associated probabilities that these changes are true. This representation is different from the conventional change detection result representation, i.e., a receiver operating characteristic (ROC) curve showing the relationship between the average detection probability and the false alarm rate.