Anomaly localization plays a critical role if a disaster occurs in a high-density crowd to efficiently rescue the crowd from the right location. This paper enriches anomaly localization by introducing new localization features, in addition to the �only� localization feature of source/start point detection existing literature offers. New features introduced include crowd density estimation in localized regions and direction/angle of a localized region. A motion shape image (MSI) based approach is introduced for localization and features detection and experimentation is performed on benchmark and our proposed high-density crowd datasets. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.