Development of a Device for Assessing the Emotional State of Companion Dogs Based on the Analysis of Vocalizations Associated with Anxiety and Barking

Introduction. Canine behaviour, socialization and interaction of dogs with humans are the key factors to be assessed among the risks inducing the increase of incidents related with development of anxiety in dogs. First of all, the lack of knowledge about the reasons of problem behaviour in animals and the high individual variability of behavioural patterns explain the relevance of this work. Moreover, the different methods used by the practitioners and absence of traceability of the owneranimal interaction record lead to the incorrect accumulation of diverse information about the identification and correction of companion dogs’ problem behaviour. The authors of this article aimed to develop a device for registering anxious and aggressive behaviour in dogs by analysing the intervalograms and spectrograms of vocalizations.

Materials and Methods. The recordings of canine vocalizations (250 recordings) were obtained as a result of authors' observations carried out in Rostov-on-Don in the period from spring to autumn 2021. The growth of signal intensity was detected by a noise sensor or microphone. The amplitude and duration of the signal were determined by a microcontroller. The Arduino Integrated Development Environment (IDE) was used for writing a programme and flashing the microcontroller. The timers calculating the number of milliseconds from the beginning to the end of an event were implemented in the programme, also, the counters measuring the number of the events within a certain period of time were used.

Results. A flowchart of the device for registering and classifying canine vocalizations, which are considered to be the behavioural anxiety markers, has been developed and described. The algorithm for assessing the type of animal activity during vocalization has been suggested. The device prototype has been developed, which enables determination and sending to a Telegram chat bot the data on dog’s anxiety and activity, as well as on the ambient temperature, based on the calculation and analysis of the vocalizations intervalograms.

Discussion and Conclusion. The significance of the present development lies in creation of the hardware and software complex prototype that provides the unbiased analysis of the information about changes in canine behaviour in the real time. The implementation of this complex makes it possible to assess the canine behaviour and provides the opportunity to obtain the new data on probability of behavioural disorders in animals caused by a high level of anxiety. The complex can serve as a prototype for creating the systems for tracking and identifying other animal species’ behaviour (including in the urban settings).