Automated classification of quilt photographs into crazy and non-crazy

Gokhale, Alhaad; Bajcsy, Peter

doi:10.1117/12.876370

Cited by 2 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, individual patches were traditionally often decorated with detailed, and often representational, needlework which is anything but random. Finally, the colour selection of patches appears, at least in most cases, to be non-random (although we did not analyse colour in the current study, and accurate determination of a single “colour” for the complex fabric patches typical of our quilts is far from trivial, see [26]). …”

Section: Discussionmentioning

confidence: 96%

Spatial Analysis of “Crazy Quilts”, a Class of Potentially Random Aesthetic Artefacts

Westphal-Fitch

Fitch

2013

PLoS ONE

View full text Add to dashboard Cite

Human artefacts in general are highly structured and often display ordering principles such as translational, reflectional or rotational symmetry. In contrast, human artefacts that are intended to appear random and non symmetrical are very rare. Furthermore, many studies show that humans find it extremely difficult to recognize or reproduce truly random patterns or sequences. Here, we attempt to model two-dimensional decorative spatial patterns produced by humans that show no obvious order. “Crazy quilts” represent a historically important style of quilt making that became popular in the 1870s, and lasted about 50 years. Crazy quilts are unusual because unlike most human artefacts, they are specifically intended to appear haphazard and unstructured. We evaluate the degree to which this intention was achieved by using statistical techniques of spatial point pattern analysis to compare crazy quilts with regular quilts from the same region and era and to evaluate the fit of various random distributions to these two quilt classes. We found that the two quilt categories exhibit fundamentally different spatial characteristics: The patch areas of crazy quilts derive from a continuous random distribution, while area distributions of regular quilts consist of Gaussian mixtures. These Gaussian mixtures derive from regular pattern motifs that are repeated and we suggest that such a mixture is a distinctive signature of human-made visual patterns. In contrast, the distribution found in crazy quilts is shared with many other naturally occurring spatial patterns. Centroids of patches in the two quilt classes are spaced differently and in general, crazy quilts but not regular quilts are well-fitted by a random Strauss process. These results indicate that, within the constraints of the quilt format, Victorian quilters indeed achieved their goal of generating random structures.

show abstract

Section: Discussionmentioning

confidence: 96%

Spatial Analysis of “Crazy Quilts”, a Class of Potentially Random Aesthetic Artefacts

Westphal-Fitch

Fitch

2013

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…Content-based image retrieval based on image color is one of the most widely used techniques. 47 Use of color computer vision can range from authenticating Jackson Pollack paintings 48 to distinguishing different types of quilt patterns, 49 demonstrating an ability to distinguish colors precisely. However, while projects such as these establish capability for such exact distinction, they use color as a processing tool rather than an end result.…”

Section: Discussionmentioning

confidence: 99%

Color by Numbers: An Exploration of the Use of Color as Classification Notation

Clarke

2013

Art Documentation: Journal of the Art Libraries Society of Nort

View full text Add to dashboard Cite

Notation is a fundamental component of a classification scheme, especially library and bibliographic classification. However, notation is often considered an afterthought or auxiliary to classification itself. With the advances in technology, classification systems, including their notation, must evolve. What, if any, possibilities lie beyond alphanumeric characters and symbols? The author explores the possible use of color as classificatory notation by looking at the traditional qualities of notation and the classificatory needs it must accommodate, various theories and standards of color, and their possible applications to classification notation. Theoretical and practical implications are considered and discussed, as well as larger implications for notation and classification overall. introduction Color and classification intertwine throughout history. Classification consists of ordering subjects and/or objects into groups called "classes" based on characteristics of division-i.e., things they have in common or things that make them different. Color is an obvious physical characteristic by which to group like items. Elaborate color classification schemes arose over time to classify things like insects, soil, 1 and even human races based on skin tone. 2 It is obvious that there is a fundamental relationship between color and classification. In addition to using color as a classificatory characteristic of division, tints and hues themselves are organized into standardized color schemes, such as the Pantone system 3 and assigned identifying notational labels based on pigment mixtures and combinations. All of these examples reflect color

show abstract

Automated classification of quilt photographs into crazy and non-crazy

Cited by 2 publications

References 6 publications

Spatial Analysis of “Crazy Quilts”, a Class of Potentially Random Aesthetic Artefacts

Spatial Analysis of “Crazy Quilts”, a Class of Potentially Random Aesthetic Artefacts

Color by Numbers: An Exploration of the Use of Color as Classification Notation

Contact Info

Product

Resources

About