Disocclusion Mitigation For Point Cloud Image-Based Level-Of-Detail Imposters