In recent years, "frailty," which describes a vulnerable condition, has been attracting attention in the approach to extend healthy life expectancy. The main factors of frailty include muscle weakness and hypoactivity. The purpose of this study is to develop a healthcare system that can easily estimate lower limb muscle strength from activities of daily living for strengthening that muscle and promoting behavioral changes. We focused on the relation of daily walking motion and lower limb muscle strength and selected two effective features from the accuracy of the estimation model using regression analysis, then verified some systems based on different machine learning models with these features. As a result, an estimation system was created and it became possible to construct a basic framework for understanding the state of lower limb muscle strength.