Umeå University's logo

Zoos, as “hotspots” for tourists and visitors, are making positive economic contributions to many countries like Sweden and Australia. In this paper, we investigate how to enrich people's zoo visiting experience with a wireless sensor network and in particular with “participatory computing” of the zoo visitors. We proposed a conceptual framework with simulation, where the zoo visitors are allowed to contribute their discoveries and captured information to the wireless sensor network via their mobile devices. The visitors' supplied information is subsequently processed to help to find interesting animals or animal events in the zoo and also to enable a number of interesting services to the visitors. In particular, the visitors' contributions are divided into different levels according to the amount of efforts required by each participatory task. In return to visitors' contributions, they are subsequently offered various services in different levels too such as animal searching, augmented reality service on mobiles, etc.

Knowledge about the indoor occupancy is one of the important sources of information to design smart buildings. In some applications, the number of occupants in each zone is required. However, there are many challenges such as user privacy, communication limit, and sensor’s computational capability in development of the occupancy monitoring systems. In this work, a people flow counting algorithm has been developed which uses low-resolution thermal images to avoid any privacy concern. Moreover, the proposed scheme is designed to be applicable for wireless sensor networks based on the internet-of-things platform. Simple low-complexity image processing techniques are considered to detect possible objects in sensor’s field of view. To tackle the noisy detection measurements, a multi-Bernoulli target tracking approach is used to track and finally to count the number of people passing the area of interest in different directions. Based on the sensor node’s processing capability, one can consider either a centralized or a full in situ people flow counting system. By performing the tracking part either in sensor node or in a fusion center, there would be a trade off between the computational complexity and the transmission rate. Therefore, the developed system can be performed in a wide range of applications with different processing and transmission constraints. The accuracy and robustness of the proposed method are also evaluated with real measurements from different conducted trials and open-source dataset.

There is a need for reliable and efficient methods for monitoring the activity and social behaviour in cows, in order to optimise management in modern dairy farms. This research presents an embedded system that could track individual cows using Ultra-wideband technology. At the same time, social interactions between individuals around the feeding area were analysed with a computer vision module. Detections of the dairy cows' negative and positive interactions were performed on foreground video stream using a Long-term Recurrent Convolution Networks model. The sensor fusion system was implemented and tested on seven dairy cows during 45 days in an experimental dairy farm. The system performance was evaluated at the feeding area. The real-time locating system based on Ultra-wideband technology reached an accuracy with mean error 0.39 m and standard deviation 0.62 m. The accuracy of detecting the affiliative and agonistic social interactions reached 93.2%. This study demonstrates a potential system for monitoring social interactions between dairy cows.

The growing interest in precision livestock farming is prompted by a desire to understand the basic behavioural needs of the animals and optimize the contribution of each animal. The aim of this study was to develop a system that automatically generated individual animal behaviour and localization data in sheep. A sensor-fusion-system tracking individual sheep position and detecting sheep standing/lying behaviour was proposed. The mean error and standard deviation of sheep position performed by the ultra-wideband location system was 0.357 +/- 0.254 m, and the sensitivity of the sheep standing and lying detection performed by infrared radiation cameras and three-dimenional computer vision technology were 98.16% and 100%, respectively. The proposed system was able to generate individual animal activity reports and the real-time detection was achieved. The system can increase the convenience for animal behaviour studies and monitoring of animal welfare in the production environment.

Collisions between vehicles and animals have increased in Norway and Europe over the last 40 years, causing economical losses as well as poor welfare for animals and people. In Norway 2 m high deer fences have been put up to prevent killings of semi-domestic reindeer at some limited road- and railway distances. However, these are very expensive and animal passages are needed. As a supplementary measure, an electronic system, aimed at warning the driver if reindeer are close to the road, was tested along the E6 main road at Saltfjellet, Norway, during March - April 2018. 235 female reindeer were equipped with radio transmitter collars (805.15.4 866 MHz). The animals were grazing in a collision exposed area at Saltfjellet. A total of 41 receivers were mounted on road sticks alongside the 4.5 km test distance. When a reindeer with transmitter was within 50 - 100 m proximity to a road stick, the receiver started blinking red. Saltfjellet reindeer herding district lost 15 reindeer in the test area from December 2017 until test start in February 2018. No reindeer were hit by cars during the test period. Nevertheless, 25% of the receivers became defective after a while, most possibly due to battery shortage. A new generation of transmitters and receivers was tested at Saltfjellet during December 2018 - April 2019. Preliminary results from this extended test are presented. We evaluate the electronic warning system as promising. However, improvements are needed in order to achieve optimal receiver and transmitter reliability in function.

The growing interest of animal welfare is prompted amongst other by understanding basic behavioural need of the animals. The aim of this study was to develop a system that automatically generates animal activity data. Therefore, a computer vision-based system for detecting sheep standing and lying behaviour was proposed. The system was composed of a multi-camera video recording system and a software module which can detect sheep standing/ lying behaviour by using the depth video stream and infrared video stream. Assessment of the detection results were carried out by comparison with the results by observation. The sensitivity of the system achieved for detecting sheep standing and lying was 96.4% and 94.16% respectively. The proposed system was able to compute sheep behaviour and the real-time detection can be achieved. The system can increase the convenience for animal behaviour studies and monitoring of animal welfare in the production environment.

This paper purposes an application based on video supervision systems in the zoo for human animal computer interaction. Bear-Watcher system covers the entire process from collecting animal's visual data, analyzing their movement and behavior, and presenting them to user interface for tourism and animal welfare. With the interaction between the users and animal movement information, the system gives the tourist more digital, more available, more involved experience. In the meanwhile, zoo keepers get reliable, accurate, cost-effective way to take care of animals.

In this paper we will describe a positioning system based on synchronized IR light. Each node will be assigned a timeslot where they will send out an IR light. There is an IR camera that is also synchronized to the timeslots that will detect the position of each node and the ID that corresponds to the timeslot. To synchronize the clock of all nodes an IR flashlight is sent out that is detected by a photodiode on the nodes. The demo will show live video stream from a network camera where the ID and position of each node in view will be overlaid in real-time in the video.

In Norway, Sweden and Finland approximately 9.000 reindeer are killed every year in traffic accidents (road and railway). In field experiments two different electronically sensor system were used. The purpose is to warn car drivers and railway engineers that reindeer are close to, or on the traffic system, and reduce the numbers of accident. The information system developed by Telespor is based on GPS/GSM radio collars. These collars use SMS to inform that the reindeer are close to the traffic system. In addition to the warning, the reindeer owners get information about the reindeer land use throughout the whole year. The collars are expensive (approx. 250 euros/unit) and need battery change every year. At Umeå University a system based on wireless sensor nodes attached to the reindeer collars is developed. This node will send out a periodic warning radio beacon. On road sticks a warning device will be attached containing a radio receiver and LED flashlights that will warn if a radio beacon is received. To further increase the warning distance it is possible to use communication between the roadside warning devices and let several devices flash their LED lights when a reindeer is detected. The radio collar (approx. 15 euros/unit) will have a theoretical battery life of more than 10 years. The radio range is around 500 meters between the radio collar and the roadside warning device, which is more than sufficient.

In this paper the authors present an approach to provide efficient low-complexity encoding for the block-based video coding scheme. The authors present a method based on removing the most time-consuming task, that is motion estimation, from the encoder. Instead the decoder will perform motion prediction based on the available decoded frame and send the predicted motion vectors to the encoder. The results presented are based on a modified H.264 implementation. The results show that this approach can provide rather good coding efficiency even for relatively high network delays.