Advanced search


Knowledge area




4 results, page 1 of 1

Color pattern and body size variation in live Aspidoscelis costatus costatus (Squamata: Teiidae) from a protected enclave in southern Mexico

ALDO GOMEZ BENITEZ Oswaldo Hernández Gallegos Brittany Lovell Pelagie Kadia JAMES MARTIN WALKER (2020)

Coloración en la lagartija Aspidoscelis costatus costatus

Whiptail lizards in the sexlineatus species group (genus Aspidoscelis) in North America represent some of the most challenging patterns of variation in the North American herpetofauna. The range of color patterns in these populations is based on individual, ontogenetic, sexual, seasonal, and/or geographic variation. We studied representatives of a population of Western Mexico Whiptail (A. costatus costatus) from a protected private enclave of approximately 0.27 ha in the municipality and city of Ixtapan de la Sal, Estado de México, México. We captured 50 lizards in 2016 and 24 in 2018, most of which we photographed ex situ and a few in situ. These photographs revealed that a variety of age/size related dorsal and ventral patterns were consistently present. Males progressed through five stages of dorsal pattern changes from pale stripes, dark intervening fields, no spots to spots, and diverse pale configurations set in a black ground color. Females in this population showed similar changes but did not lose striping as they grew. Ontogenetic changes in ventral color patterns were also apparent, with males becoming more colorful than females. The adaptive significance of extensive color pattern variation in this urban population of A. c. costatus warrants further study.

Article

adaptive significance Balsas Basin Whiptail body size dorsal coloration Mexican lizards ontogeny spots stripes ventral coloration BIOLOGÍA Y QUÍMICA

PERCEPTIONS OF CURRICULAR CHANGES AND APPROPRIATION OF NEW EDUCATIONAL PRACTICES FOR THE LEARNING OF ENGLISH AS A FOREIGN LANGUAGE IN PUBLIC EDUCATION: A MIXED METHODS STUDY WITH TELESECONDARY SCHOOL TEACHERS IN TABASCO A DISSERTATION

MIGUEL HERNANDEZ HERNANDEZ (2020)

In Mexico, as in other parts of the globe, educational reforms are sanctioning curricular changes that favour the learning of English as a Foreign Language (EFL) in public education (Izquierdo, García, Garza, & Aquino, 2016). To date, research has examined how public education EFL teachers in urban schools perceived these changes or the extent to which they adopted the suggested teaching practices (Altaieb, 2013; Borg, 2006). Nonetheless, EFL policies have reached rural schools where “generalist” teachers deliver English instruction along with all areas of the curriculum (SEP, 2017). This study explores perceptions about EFL curricular changes and the appropriation of the suggested teaching practices among generalist rural secondary school teachers in Tabasco, Mexico. To this end, data were collected from 216 secondary-school teachers from 11 telesecondary schools through an explanatory sequential mixed methods design (Creswell &Creswell, 2018). The quantitative data were collected using two Likert-scale questionnaires that elicited information about perceptions of curricular changes and appropriation of new practices. Qualitative data were collected through a 40-minute semi-structured interview, which was administered to a sub-sample of 14 participants who exhibited either very positive or negative answer tendencies in both questionnaires. The quantitative data were analyzed using SPSS v.25, first, to determine the Cronbach alpha reliability coefficients of each dimension of both Likert-scale questionnaires. Then, Spearman correlation analyses were used to examine the correlation within and between the perception and the appropriation questionnaires.

276

Doctoral thesis

HUMANIDADES Y CIENCIAS DE LA CONDUCTA CHANGES AND APPROPRIATION EDUCATIONAL PRACTICES TELESECONDARY SCHOOL TABASCO

Reconocimiento de patrones invariante a desplazamiento, escala y rotación vía transformada radial de Hilbert optimizada

Displacement, scale and rotation invariant pattern recognition by Radial Hilbert Transform Optimized

José Alfredo Castro Valdez (2021)

La identificación y clasificación de imágenes tiene múltiples aplicaciones, así como metodologías para llevarse a cabo. En este trabajo se utilizó una metodología con invariancia a desplazamiento, escala y rotación, estas invariancias se logran al utilizar un conjunto de transformadas. La transformada de Fourier tiene la propiedad de invarianza a desplazamiento mientras que la transformada de Mellin genera la invarianza a escala, por otra parte, la transformada radial de Hilbert transforma una imagen en una firma vectorial, la cual tiene invarianza a rotación. Esta nueva metodología toma una imagen centrada y le aplica la trasformada Fourier-Mellin. Se utilizan las máscaras de anillos concéntricos obtenidos a partir de la transformada Radial de Hilbert y se hace una multiplicación punto a punto con el módulo de la transformada FourierMellin. Los valores en cada anillo se suman y con ellos se forma una firma vectorial cuya longitud está relacionada con el número de anillos concéntricos y además es invariante a escala, rotación y desplazamiento. En este trabajo se partió de la metodología previamente mencionada y se propuso una adaptación de la transformada radial de Hilbert en el cual la cantidad de anillos concéntricos obtenida fuera máxima, a esta versión se le llamó transformada radial de Hilbert optimizada (RHTO). Para comparar entre firmas vectoriales se utilizó el filtro clásico y se desarrolló un nuevo filtro de correlación que se adapta entre los casos lineal y no lineal dependiendo de firmas a correlacionar, a este filtro se le llamó filtro lineal / no-lineal adaptativo. Para evaluar la metodología propuesta se utilizaron imágenes de 30 especies de fitoplancton con diferentes escalas y rotaciones. Además, a cada combinación de escala y rotación se les aplicó un conjunto patrones de iluminación no homogéneo obteniendo. Los valores de correlación obtenidos fueron analizados con una matriz de confusión donde se compararon diferentes parámetros estadísticos, encontrando que al utilizar la metodología RHTO y el filtro adaptativo se obtienen mejores valores de clasificación exitosa.

Image classification has multiple applications and methodologies to be carried out. In this work, a method with invariance to displacement, scale, and rotation was used. These invariances are achieved by using a set of transforms. The Fourier transform has the property of displacement invariance, while the Mellin transform generates scale invariance. On the other hand, the radial Hilbert transform transforms an image into a vector signature, which has rotation invariance. This new methodology takes a centered image and applies the Fourier-Mellin transform to it. The concentric ring masks obtained from the Radial Hilbert transform are used, and a point-to-point multiplication is made with the module of the Fourier-Mellin transform. The values in each ring are added, and with them, a vector signature is formed whose length is related to the number of concentric rings and is also invariant to displacement, scale, and rotation. In this work, we started from the previously mentioned methodology, and an adaptation of the radial Hilbert transform was proposed in which the number of concentric rings obtained was maximum. This version was called optimized radial Hilbert transform (RHTO). When correlating images in one or two dimensions, you must decide the correlation filter to use. The new methodology used the classical filter. A new filter was developed that adapts between linear and nonlinear cases depending on the signatures correlated; this filter was called the adaptive linear/nonlinear filter. Thirty phytoplankton species with different scales and rotations were used to evaluate the proposed methodology. In addition, a set of non-homogeneous lighting patterns was applied to each combination of scale and rotation. A confusion matrix analyzed the correlation values obtained where different statistical parameters were compared, finding that better values for successful classification were obtained when using the RHTO methodology and the adaptive filter.

Doctoral thesis

Transformada radial de Hilbert optimizada, correlación adaptativa, Firmas unidimensionales, Transformada de Fourier, Reconocimiento de patrones, Clasificación de imágenes Radial Hilbert transform optimized, Adaptive correlation, Unidimensional signatures, Fourier transform, Pattern recognition, Image classification CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA FÍSICA ÓPTICA OPTICA FÍSICA OPTICA FÍSICA