This paper presents a probabilistic approach for the detection of abnormal behaviour in Activities of Daily Living (ADLs) from sensor data collected from 30 participants. The ADLs considered are: (i) preparing and drinking tea, and (ii) preparing and drinking coffee. Abnormal behaviour detected in the context of these activities can be an indicator of a progressive health problem or the occurrence of a hazardous incident. The approach presented considers the temporal aspect of the sequences of actions that are part of each ADL and that vary between participants. The average and standard deviation for the durations of each action were calculated to define an average time and a range in which a behaviour could be considered as normal for each stage and activity. The Cumulative Distribution Function (CDF) was used to obtain the probabilities of abnormal behaviours related to the early and late completion of activities and stages within an activity. The data analysis show that CDF can provide accurate and reliable results regarding the presence of abnormal behaviour in stages and activities that last over a minute. The approach presented could be used to train and improve machine learning algorithms of abnormal behaviour detection.

This paper presents a probabilistic approach for the identification of abnormal behaviour in Activities of Daily Living (ADLs) from dense sensor data collected from 30 participants. The ADLs considered are related to preparing and drinking (i) tea, and (ii) coffee. Abnormal behaviour identified in the context of these activities can be an indicator of a progressive health problem or the occurrence of a hazardous incident. The approach presented considers the temporal and sequential aspects of the actions that are part of each ADL and that vary between participants. The average and standard deviation for the duration and number of steps of each activity are calculated to define the average time and steps and a range within which a behaviour could be considered as normal for each stage and activity. The Cumulative Distribution Function (CDF) is used to obtain the probabilities of abnormal behaviours related to the early and late completion of activities and stages within an activity in terms of time and steps. Analysis shows that CDF can provide precise and reliable results regarding the presence of abnormal behaviour in stages and activities that last over a minute or consist of many steps. Finally, this approach could be used to train machine learning algorithms for abnormal behaviour detection.

introduction of a new health-care delivery paradigm [3]. Besides, low-cost sensing solutions, whose wireless services coupled with rapid advances in data analysis, have provided the next generation of products to be deployed within living environments. These have the potential to improve the manner where remote health-care support can be provided and are slowly gaining increased acceptance by both users and health-care professionals [4]. From the multitude of health scenarios to consider, detecting falls within the living environment is a relevant challenge with a high impact in terms of both security and safety. Accidental falls can cause serious injury to at-risk individuals, especially for the aging [5]. Within this cohort, falls are the leading cause of hospitalization, injury-related deaths and loss of independence. However, it has been demonstrated that detecting and rapidly responding to falls can reduce the long-term risks associated with falls. Although efforts have been directed towards supporting the detection and management of falls within living environments, a range of issues still exist. From a usability perspective, challenges are faced by the costs of the solution and the perceived issue of intrusiveness when video based cameras are used. From a technical perspective, challenges are faced by levels of accuracy levels and a desire to reduce the numbers of false positives given the implications that these have from a health-care provision perspective. In addition, the studies of fall detection are mainly

This paper presents a comparison between algorithms (Oriented FAST and Rotated BRIEF (ORB) andAruco) for the detection of fiducial markers placed throughout a smart environment. A series of activities of daily living (ADL) were conducted while monitoring a first-person perspective of the situation; this was achieved through the usage of the Google Glass platform. Fiducial markers were employed, as a means to assist with the detection of specific objects of interest, within the environment. Each marker was assigned unique Identification (ID) and was used to identify the object. Three activities were performed by a participant within the environment. On subsequent trials of the solution, lighting conditions were modified to assess fiducial marker detection rates on a frame-by-frame basis. This paper presents the results from this investigation, detailing performance measure for each object detected under various lighting conditions, motion blur and distance from the objects. An intelligent system was developed to specifically consider distance estimation in order to aid with the filtering out of false interactions. A linear filtering method was applied along with a fuzzy membership function to estimate the degree of user interaction, which assists in removing false positives generated by the occupant. The intelligent system returns an average precision, recall and an F-Measure of 0.99, 0.62 and 0.49, respectively.

Gait velocity is an important measure of independence and functional ability to those within the older population. Detecting changes in gait velocity can aid to provide interventions to avoid hospitalisation, currently gait velocity is assessed in a clinical setting, where the patient is timed over a measured distance between 3-6 metres by a clinician, however, this is time consuming, subjective, and not possible to carry out frequently over time. An unobtrusive method of monitoring gait velocity, frequently, over extended periods of time, would therefore be advantageous when developing interventions. This paper proposes an unobtrusive computer vision-based method of continuously monitoring an occupants gait velocity within their own home. This is achieved through the use of a low cost thermal vision sensor. The system was benchmarked against the clinical standard method of being timed by a stopwatch. Results show a high correlation between the gait velocity measured by the thermal vision sensor and the measured stopwatch velocity (R=0.941, p=0.02).