Human figure segmentation using independent component analysis

In this paper, we present a Statistical Shape Model for Human Figure Segmentation in gait sequences. Point Distribution Models (PDM) generally use Principal Component analysis (PCA) to describe the main directions of variation in the training set. However, PCA assumes a number of restrictions on the data that do not always hold. In this work, we explore the potential of Independent Component Analysis (ICA) as an alternative shape decomposition to the PDM-based Human Figure Segmentation. The shape model obtained enables accurate estimation of human figures despite segmentation errors in the input silhouettes and has really good convergence qualities.

2D silhouette and 3D skeletal models for human detection and tracking

In this paper we propose a statistical model for detection and tracking of human silhouette and the corresponding 3D skeletal structure in gait sequences. We follow a point distribution model (PDM) approach using a Principal Component Analysis (PCA). The problem of non-lineal PCA is partially resolved by applying a different PDM depending of pose estimation; frontal, lateral and diagonal, estimated by Fisher's linear discriminant. Additionally, the fitting is carried out by selecting the closest allowable shape from the training set by means of a nearest neighbor classifier. To improve the performance of the model we develop a human gait analysis to take into account temporal dynamic to track the human body. The incorporation of temporal constraints on the model increase reliability and robustness.

Newly Acquired Fear of Falling Leads to Altered Eye Movement Patterns and Reduced Stepping Safety: A Case Study

This opportune case study describes visual and stepping behaviours of an 87 year old female (P8), both prior to, and following two falls. Before falling, when asked to walk along a path containing two stepping guides positioned before and after an obstacle, P8 generally visually fixated the first stepping guide until after foot contact inside it. However, after falling P8 consistently looked away from the stepping guide before completing the step into it in order to fixate the upcoming obstacle in her path. The timing of gaze redirection away from the target (in relation to foot contact inside it) correlated with absolute stepping error. No differences in eyesight, cognitive function, or balance were found between pre- and post-fall recordings. However, P8 did report large increases in fall-related anxiety and reduced balance confidence, supporting previously suggested links between anxiety/increased fear or falling and maladaptive visual/stepping behaviours. The results represent a novel insight into how psychological and related behavioural factors can change in older adults following a fall, and provide a possible partial rationalisation for why recent fallers are more likely to fall again in the following 12 months. These findings highlight novel possibilities for falls prevention and rehabilitation.

The influence of providing perches and string on activity levels, fearfulness and leg health in commercial broiler chickens

The aim of this study was to assess the effect of providing environmental enrichment in the form of perches and string on the behaviour and welfare of commercial broiler chickens. Houses containing ~23 000 broiler chickens were assigned to one of four treatments in a 2×2 factorial design. Treatments involved two levels of access to perches (P) (present (24/house) ‘+P’ or absent ‘−P’) and two levels of access to string (S) (present (24/house) ‘+S’ or absent ‘−S’). All houses contained windows, and 30 straw bales were provided from day 10 of the rearing cycle. Treatments were applied in one of four houses on a single farm, and were replicated over four production cycles. Behaviour and leg health were observed in weeks 3 to 5 of the rearing cycle. Production performance and environmental parameters were also measured. There was an interaction between perches and age in the percentage of birds observed lying, with higher percentages of birds observed lying in the +P treatment than in the −P treatment during weeks 4 and 5. There was also a significant interaction between string and age in the percentage of birds observed in locomotion, with higher percentages observed in locomotion in the −S treatment than in the +S treatment during weeks 4 and 5. There was also an interaction between string and age in average gait scores, with lower gait scores in the +S treatment than in the −S treatment during weeks 3 and 5 but not within week 4. Daytime observations showed that perches and strings were used frequently, with one bout of perching occurring approximately every 80 s/perch, and one bout of pecking at string occurring every 78 s/string on average. There was a significant effect of age on use of perches (P<0.001) and string (P<0.001), with perching peaking during week 5 and string pecking peaking during week 3. We conclude that commercial broilers in windowed houses with access to straw bales display an interest in additional enrichment stimuli in the form of perches and string, and therefore that these stimuli have the potential to improve welfare. In addition, provision of string as a pecking device appeared to positively influence walking ability. However, this effect was numerically small, was only shown in certain weeks and was not reflected in the other leg health measure (latency to lie). The results also showed an apparent negative effect of string and perches on the activity levels of birds (recorded away from the immediate vicinity of these enrichments) towards the end of the production cycle. These results emphasise the need for further research into optimum design and layout of enrichment stimuli for modern broilers in windowed houses to ensure that their provision leads to clear welfare benefits.