Limits...
Visual perception of procedural textures: identifying perceptual dimensions and predicting generation models.

Liu J, Dong J, Cai X, Qi L, Chantler M - PLoS ONE (2015)

Bottom Line: The results suggested that existing dimensions in literature cannot accommodate random textures.We therefore utilized isometric feature mapping (Isomap) to establish a three-dimensional perceptual texture space which better explains the features used by humans in texture similarity judgment.Finally, we proposed computational models to map perceptual features to the perceptual texture space, which can suggest a procedural model to produce textures according to user-defined perceptual scales.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science and Technology, Ocean University of China, 238 Songling Road, Qingdao, Shandong, China; Science and Information College, Qingdao Agricultural University, 700 Changcheng Road, Qingdao, Shandong, China.

ABSTRACT
Procedural models are widely used in computer graphics for generating realistic, natural-looking textures. However, these mathematical models are not perceptually meaningful, whereas the users, such as artists and designers, would prefer to make descriptions using intuitive and perceptual characteristics like "repetitive," "directional," "structured," and so on. To make up for this gap, we investigated the perceptual dimensions of textures generated by a collection of procedural models. Two psychophysical experiments were conducted: free-grouping and rating. We applied Hierarchical Cluster Analysis (HCA) and Singular Value Decomposition (SVD) to discover the perceptual features used by the observers in grouping similar textures. The results suggested that existing dimensions in literature cannot accommodate random textures. We therefore utilized isometric feature mapping (Isomap) to establish a three-dimensional perceptual texture space which better explains the features used by humans in texture similarity judgment. Finally, we proposed computational models to map perceptual features to the perceptual texture space, which can suggest a procedural model to produce textures according to user-defined perceptual scales.

No MeSH data available.


The resulting dendrogram of HCA.Three clusters below the dissimilarity level of 7 were labeled as Cluster A, Cluster B and Cluster C. Models which were classified as groups below the dissimilarity level of 2.5 were represented by different colors.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4481328&req=5

pone.0130335.g003: The resulting dendrogram of HCA.Three clusters below the dissimilarity level of 7 were labeled as Cluster A, Cluster B and Cluster C. Models which were classified as groups below the dissimilarity level of 2.5 were represented by different colors.

Mentions: The resulting dendrogram illustrated in Fig 3 showed how the texture generation models were clustered into different groups. In Fig 3, there were three chunks below a dissimilarity level of 7. Models of “Matrix Transformation”, “Texton Regular” and “Islamic Patterns” were clustered into one group; we named this chunk as Cluster A. However, the “Matrix Transformation” model was separated from the other two models, which meant that textures produced by the “Matrix Transformation” were most dissimilar to others. In fact, the “Matrix Transformation” was a unique method that was capable of generating fabric-like textures which were perceived as uniform, locally oriented and regular. Texture patterns produced by “Islamic Patterns” and “Texton Regular” can be described as regular, directional, and repetitive. In these textures, structural primitives were distributed repetitively and regularly. Cluster B, comprising of 11 models, generated textures with granular textons randomly spread in the images. Tactile roughness appeared in textures generated by models in Cluster B. Cluster C comprised of 9 models, and textures in this category consisted of randomly distributed near-regular shape elements. They shared the features of repetition, near regularity, roughness and coarseness. In this cluster, the “Fractal (one-over-fBeta-noise)” model had large dissimilarity with others, and textures appeared as noticeably vertical stripes. Moreover, models clustered as groups below the dissimilarity level of 2.5 were regarded as being able to generate textures resembling each other. These were marked with different colors in Fig 3.


Visual perception of procedural textures: identifying perceptual dimensions and predicting generation models.

Liu J, Dong J, Cai X, Qi L, Chantler M - PLoS ONE (2015)

The resulting dendrogram of HCA.Three clusters below the dissimilarity level of 7 were labeled as Cluster A, Cluster B and Cluster C. Models which were classified as groups below the dissimilarity level of 2.5 were represented by different colors.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4481328&req=5

pone.0130335.g003: The resulting dendrogram of HCA.Three clusters below the dissimilarity level of 7 were labeled as Cluster A, Cluster B and Cluster C. Models which were classified as groups below the dissimilarity level of 2.5 were represented by different colors.
Mentions: The resulting dendrogram illustrated in Fig 3 showed how the texture generation models were clustered into different groups. In Fig 3, there were three chunks below a dissimilarity level of 7. Models of “Matrix Transformation”, “Texton Regular” and “Islamic Patterns” were clustered into one group; we named this chunk as Cluster A. However, the “Matrix Transformation” model was separated from the other two models, which meant that textures produced by the “Matrix Transformation” were most dissimilar to others. In fact, the “Matrix Transformation” was a unique method that was capable of generating fabric-like textures which were perceived as uniform, locally oriented and regular. Texture patterns produced by “Islamic Patterns” and “Texton Regular” can be described as regular, directional, and repetitive. In these textures, structural primitives were distributed repetitively and regularly. Cluster B, comprising of 11 models, generated textures with granular textons randomly spread in the images. Tactile roughness appeared in textures generated by models in Cluster B. Cluster C comprised of 9 models, and textures in this category consisted of randomly distributed near-regular shape elements. They shared the features of repetition, near regularity, roughness and coarseness. In this cluster, the “Fractal (one-over-fBeta-noise)” model had large dissimilarity with others, and textures appeared as noticeably vertical stripes. Moreover, models clustered as groups below the dissimilarity level of 2.5 were regarded as being able to generate textures resembling each other. These were marked with different colors in Fig 3.

Bottom Line: The results suggested that existing dimensions in literature cannot accommodate random textures.We therefore utilized isometric feature mapping (Isomap) to establish a three-dimensional perceptual texture space which better explains the features used by humans in texture similarity judgment.Finally, we proposed computational models to map perceptual features to the perceptual texture space, which can suggest a procedural model to produce textures according to user-defined perceptual scales.

View Article: PubMed Central - PubMed

Affiliation: Department of Computer Science and Technology, Ocean University of China, 238 Songling Road, Qingdao, Shandong, China; Science and Information College, Qingdao Agricultural University, 700 Changcheng Road, Qingdao, Shandong, China.

ABSTRACT
Procedural models are widely used in computer graphics for generating realistic, natural-looking textures. However, these mathematical models are not perceptually meaningful, whereas the users, such as artists and designers, would prefer to make descriptions using intuitive and perceptual characteristics like "repetitive," "directional," "structured," and so on. To make up for this gap, we investigated the perceptual dimensions of textures generated by a collection of procedural models. Two psychophysical experiments were conducted: free-grouping and rating. We applied Hierarchical Cluster Analysis (HCA) and Singular Value Decomposition (SVD) to discover the perceptual features used by the observers in grouping similar textures. The results suggested that existing dimensions in literature cannot accommodate random textures. We therefore utilized isometric feature mapping (Isomap) to establish a three-dimensional perceptual texture space which better explains the features used by humans in texture similarity judgment. Finally, we proposed computational models to map perceptual features to the perceptual texture space, which can suggest a procedural model to produce textures according to user-defined perceptual scales.

No MeSH data available.