Anthropometric Landmarks Extraction and Dimensions Measurement Based on ResNet

Anthropometric dimensions can be acquired in 2D images by landmarks. Body shape variance causes low accuracy and bad robustness of landmarks extracted, and it is difficult to determine the position of axis division point when dimensions are calculated by the ellipse model. In this paper, landmarks are extracted from images by convolutional neural network instead of the gradient of body outline. A general multi-ellipse model is proposed, the anthropometric dimensions are obtained from the length of different elliptical segments and the position of axis division point is determined by thickness–width ratio of body parts. Finally, an evaluation is completed based on 87 subjects, in which it turns out that the average accuracy of our method for identifying landmarks is 96.6%, when the number of rotation angles is 2, the three main dimensional errors calculated by our model are smaller than existing method, and the errors of other dimensions are also within the margin of error for garment measuring.

Quantitative heterodonty in Crocodylia: assessing size and shape across modern and extinct taxa

Heterodonty in Crocodylia and closely related taxa has not been defined quantitatively, as the teeth rarely have been measured. This has resulted in a range of qualitative descriptors, with little consensus on the condition of dental morphology in the clade. The purpose of this study is to present a method for the quantification of both size- and shape-heterodonty in members of Crocodylia. Data were collected from dry skeletal and fossil specimens of 34 crown crocodylians and one crocodyliform, resulting in 21 species total. Digital photographs were taken of each tooth and the skull, and the margins of both were converted into landmarks and semilandmarks. We expressed heterodonty through Foote’s morphological disparity, and a principal components analysis quantified shape variance. All specimens sampled were heterodont to varying degrees, with the majority of the shape variance represented by a ‘caniniform’ to ‘molariform’ transition. Heterodonty varied significantly between positions; size undulated whereas shape was significantly linear from mesial to distal. Size and shape appeared to be primarily decoupled. Skull shape correlated significantly with tooth shape. High size-heterodonty often correlated with relatively large caniniform teeth, reflecting a prioritization of securing prey. Large, highly molariform, distal teeth may be a consequence of high-frequency durophagy combined with prey size. The slender-snouted skull shape correlated with a caniniform arcade with low heterodonty. This was reminiscent of other underwater-feeding tetrapods, as they often focus on small prey that requires minimal processing. Several extinct taxa were very molariform, which was associated with low heterodonty. The terrestrial peirosaurid shared similarities with large modern crocodylian taxa, but may have processed prey differently. Disparity measures can be inflated or deflated if certain teeth are absent from the tooth row, and regression analysis may not best apply to strongly slender-snouted taxa. Nevertheless, when these methods are used in tandem they can give a complete picture of crocodylian heterodonty. Future researchers may apply our proposed method to most crocodylian specimens with an intact enough tooth row regardless of age, species, or rearing conditions, as this will add rigor to many life history studies of the clade.

Geometric morphometrics of functionally distinct floral organs in Iris pumila: Analyzing patterns of symmetric and asymmetric shape variations

The Iris flower is a complex morphological structure composed of two trimerous whorls of functionally distinct petaloid organs (the falls and the standards), one whorl of the stamens and one tricarpellary gynoecium. The petal-like style arms of the carpels are banded over the basal part of the falls, forming three pollination tunnels, each of which is perceived by the Iris pollinators as a single bilaterally symmetrical flower. Apart from the stamens, all petaloid floral organs are preferentially involved in advertising rewards to potential pollinators. Here we used the methods of geometric morphometrics to explore the shape variation in falls, standards and style arms of the Iris pumila flowers and to disentangle the symmetric and the asymmetric component of the total shape variance. Our results show that symmetric variation contributes mostly to the total shape variance in each of the three floral organs. Fluctuating asymmetry (FA) was the dominant component of the asymmetric shape variation in the falls and the standards, but appeared to be marginally significant in the style arms. The values of FA indexes for the shape of falls (insects? landing platforms) and for the shape of standards (long-distance reward signals) were found to be two orders of magnitude greater compared to that of the style arms. Directional asymmetry appeared to be very low, but highly statistically significant for all analyzed floral organs. Because floral symmetry can reliably indicate the presence of floral rewards, an almost perfect symmetry recorded for the style arm shape might be the outcome of pollinator preferences for symmetrical pollination units.

Transmission characteristics of a TEM waveguide for transient signals by the use of a damped sinusoidal

This article broaches the issue of the propagation of transient signals in gigahertz transverse electromagnetic (GTEM) cells. As a representative for transient signals a damped sinusoidal (DS) is used with three different mid-band frequencies. The signal transmission of the DS in the GTEM1250 is qualified and discussed on the basis of the Pearson correlation coefficient (Pcc). The Pcc gives an overview of the signal transmission quality for all measuring points within the testvolume and signal distortions can be identified. A 100 MHz DS is weakly distorted in several measuring points. The Pcc at those points decreases and a signal shape variance can be assumed. Furthermore inhomogeneities of the GTEM1250 caused by the cell door can be identified.