Freezing of Gait (FoG) is one of the most disabling symptoms in Parkinson's disease (PD). Current algorithms to detect FoG are based on wearable inertial systems which relies on the frequency response given by the inertial signal. However, these algorithms have only been evaluated under laboratory conditions causing that, in real life, they present false positives, reducing the reliability of the algorithm. This paper presents the evaluation of 20 PD patients in their homes and the inclusion of a posture algorithm in order to contextualize FoG detection. This algorithm improves the optimal FoG detection algorithm specificity from 74.5% to 79% (4.3%) while in average improves specificity from 69.9% to a 74.6% (4.7%) preserving the sensitivity. In some patients, those who performed more false positive tests, specificity could increase up to 11.95% keeping the sensitivity.