Effects of light refraction on the accuracy of camera calibration and reconstruction in underwater motion analysis

One of the most serious obstacles to accurate quantification of the underwater motion of a swimmer's body is image deformation caused by refraction. Refraction occurs at the water-air interface plane (glass) owing to the density difference. Camera calibration-reconstruction algorithms commonly used in aquatic research do not have the capability to correct this refraction-induced nonlinear image deformation and produce large reconstruction errors. The aim of this paper is to provide a thorough review of: the nature of the refraction-induced image deformation and its behaviour in underwater object-space plane reconstruction; the intrinsic shortcomings of the Direct Linear Transformation (DLT) method in underwater motion analysis; experimental conditions that interact with refraction; and alternative algorithms and strategies that can be used to improve the calibration-reconstruction accuracy. Although it is impossible to remove the refraction error completely in conventional camera calibration-reconstruction methods, it is possible to improve the accuracy to some extent by manipulating experimental conditions or calibration frame characteristics. Alternative algorithms, such as the localized DLT and the double-plane method are also available for error reduction. The ultimate solution for the refraction problem is to develop underwater camera calibration and reconstruction algorithms that have the capability to correct refraction.     

A Detection Scheme for Dependent Noise Processes

In this paper the problem of memoryless discrete time detection of a constant signal in additive φ-mixing noise is considered. The fidelity criterion is taken to be the asymptotic relative efficiency. The optimal detector nonlinearity is approximated by a polynomial, and the statistical knowledge required for the design of the detector is the family of joint moments of the noise, resulting in a very simple design procedure for the detector. It is shown that the performance of this detector can be made arbitrarily close to that of the optimal memoryless detector for a large class of noises.     

A control problem with structural choices

The case is considered in which, during the operation of an optimal control system, the optimizer, in addition to applying his usual control, may switch structures. Necessary and sufficient conditions are derived and emphasis is placed on the special characteristics of this problem. Continuous and discrete time set-ups are considered and the separation principle is shown not to hold for the linear quadratic case in the presence of noise.     

Three-axis platform simulation: Bond graph and Lagrangian approach

A bond graph model is derived for the geometric constraints of a three-axis flight table. Gimbal dynamics are easily added even in asymmetrical and unbalanced cases. A method is introduced to make the local dependent inertias computable. The bond graph compares favourably to the Lagrangian approach as to modelling effort and accessibility of intermediate variables as well as having computational advantages.     

Infant approach behavior as related to attachment.

The chief purpose of this study was to identify characteristics that distinguish approach as an attachment behavior from approach serving other behavioral systems. Locomotor approaches of 16 male and 10 female white middle-class infants to an attachment figure (the mother) and to a nonattachment figure (the visitor-observer) were examined under naturalistic conditions at home. Observations were made during 4-hour home visits at 3-week intervals; those from 26 to 54 weeks were used. When approach was examined in a free-choice situation and without regard to behavioral context, infants approached the mother proportionally (though slightly) more often than the visitor (p less than .05). In 2 behavioral contexts, however, spontaneous infant approaches were sharply differential to the mother; approaches accompanied by crying and approaches terminating in a pickup appeal were directed almost exclusively to the mother (p less than .0001 for both). Touching upon completion of the approach was not differential, except in the context of a pickup appeal. Object-oriented approaches were more often directed to the visitor than to the mother (p less than .002) and may be either exploratory or affiliative. It is suggested that neither approach nor touching can be assumed to serve the attachment system without consideration of context- both environmental and behavioral.     

The relative efficacy of vestibular-proprioceptive stimulation and the upright position in enhancing visual pursuit in neonates.

48 healthy neonates born to multiparae were randomly assigned to view a moving stimulus either in the horizontal or the upright position, with or without added vestibular stimulation and with or without pacifier sucking. The infant was shown a moving black line inside a strictly controlled visual environment provided by an apparatus which permitted horizontal and upright positioning and displacement of the infant. Visual tracking was recorded by a concealed TV camera positioned at a constant distance and angle from the infants' eyes in both positions. Quality of tracking during each of 4 trials was scored on a 7-point scale. Results indicate that vestibularproprioceptive stimulation provided to the infant either horizontally or semivertically significantly enhanced his visual tracking, whereas the upright position did not. Pacifier sucking also improved his performace.