Blur variation in 2D images caused by camera focus provides a suitable cue for depth estimation. Depth from defocus (DFD) technique calculates the blur amount in images considering that the depth and defocus blur are related to each other. Conventional DFD methods use only defocused images that might yield low-quality depth data and reconstructed infocused image. In this article, a new DFD methodology based on infocused and defocused images is proposed in which using an infocused image can solve the quality degradation problems. In this method, Subbarao’s DFD is combined with a novel edge blur estimation method to obtain an improved depth map. In addition, a saliency map mitigates the ill-posed problem of depth estimation in homogeneous regions. For real-time full high definition (Full-HD) image processing, a parallelized graphics processing unit (GPU) implementation is devised to improve execution speed.

In this article we discuss the possibility of using a conventional DSLR camera for color assessment of the prints enhanced with pearlescent pigments. Since these prints exhibit goniochromatic properties, color data were acquired in a multiangular manner and color estimation errors were assessed for the selected viewing angles. Colorimetric target-based camera characterization was performed by means of Artificial Neural Networks (ANN). In addition, ANN training was improved by implementing a multiobjective genetic algorithm with the aim to select the minimum number of different samples for the training set that will ensure efficient characterization. Our results indicate that the mean error of the performed characterization complies with the requirements placed on colorimeter in a print production. Furthermore, we show that the genetic algorithm optimization enabled an optimal training set selection for the given application, which makes the presented approach an efficient solution for multiangular color estimation.

We have previously described the Progressive Barcode, a high-density color barcode that changes over incrementally. This article expands on previously published work [S. J. Simske and A. M. Vans, “Archive-enabling tagging using progressive barcodes,” Proc. of Archiving (Los Angeles, CA, 2015), pp. 130–135]. We describe how the progressive barcode works and its applicability to information workflows and archiving applications. In particular, we tie the progressive, or sequential, aspect of this approach to the temporal relationship of multiple items (including documents in a workflow, items in a supply chain, etc.) We present an inference model that allows for the possibility of the loss of tags (called knockout) that were originally created as part of a sequence of tags, usually in manufacturing or printing as examples. Taken together, the progressive barcode with the inference model is a self-contained mechanism for the support of workflows.

Surface geometry can play an important role in our ability to understand and interpret material appearance and properties. This property ranges from large-scale shape changes impacting our identification of reflections to visible surface roughness affecting gloss perception. In this work, the authors present a user study that examines numerous surface geometries that are defined at the mesoscale: small enough to be considered indicative of the material and not object geometry, but large enough to be visible from a distance with the naked eye. Models of perceived brightness were compared against sparsely collected brightness judgments from the study and used to densely compare many generated mesoscale surface patterns. Averaging incoming luminance over a spatially varying surface proved effective at modeling brightness judgments. The effects of the mesoscale structure on perceived brightness were not directly correlated to parameters such as shape, size, or depth of the bumpy texture elements.

Visual fatigue is considered to be an indicator responsible for a user’s visual experience of watching stereoscopic display, and requires an efficient evaluation method. This article compares the visual fatigue based on electrocardiograph (ECG) analysis on 3D real objects versus stereoscopic TV, involving a magic cube based task in which participants are required to view magic cubes in the real environment and virtual images on stereoscopic TV. The results of ECG analysis show a significant increase of approximate entropy (ApEn) and stable changes of 15 other parameters. The ApEn variation shows that the visual fatigue when viewing stereoscopic TV is more severe than that when viewing magic cubes in the real environment after a 20-minute viewing session. In addition, the results for critical flicker frequency are consistent with those of ApEn, and the results of a subjective questionnaire show that the proposed experiment has induced visual fatigue. The results for correct rate also show that the visual fatigue when viewing virtual images on stereoscopic TV is more severe than that when viewing magic cubes in the real environment after a six-minute viewing session. It can be concluded that the visual fatigue when viewing stereoscopic TV appears earlier than that when viewing 3D real objects and is more severe than that when viewing 3D real objects. However, severe visual fatigue is induced both by viewing 3D real objects and by viewing stereoscopic TV after a 25-minute viewing session, and it does not show significant difference between viewing 3D real objects and viewing stereoscopic TV, which proves that the findings in this article may have potential impacts for the improvement of visual fatigue and help to guide people for comfortable viewing.

In this article, a spatial gamut mapping algorithm (SGMA) based on the guided filter is proposed. In this SGMA, the guided filter is firstly used to decompose the luminance channel of input image into high-pass signal representing image details (or textures) and low-pass signal representing image edges (or contours). The low-pass luminance signal is combined with other channels (chroma and hue) to get base-layer image. After the gamut mapping of base-layer image, the enhanced high-pass luminance signal is added back to the gamut-mapped base-layer image for detail preservation. Finally, the detail-preserved image is mapped to the destination gamut once again. In this article, the properties of the proposed SGMA, which include the principle of the guided filter, the influences of different guidance images on gamut mapping results, the influences of different frequency-band-decomposition methods on gamut mapping results, and the influences of different guided-filtering parameters on detail preservation, are analyzed in detail. The results of psychophysical experiments indicate that the proposed SGMA is better at detail preservation than pointwise GMAs, and has good performance on preference and reproduction accuracy. More importantly, the halos-testing experimental results verify the advantage of the proposed SGMA in avoiding halo-artifacts.

Based on background modeling that employs a nonlinear model, a new approach for slow moving object detection in rotating complex background is proposed. First, the rotational motion pattern is determined and the complexity of the background is evaluated. Then background modeling of rotational motion is conducted through the following three steps: (1) retrieving a set of matching feature points and obtaining feature point motion vector field; (2) filtering the vector field and obtaining precise matching feature point set; (3) estimating nonlinear model parameters and reconstructing background motion vector field. Next, background compensated image is obtained by applying pixel interpolation. Last, moving objects are detected by applying threshold segmentation and morphological processing to difference images and searching for complete object contour according to region growing rule. The main test video sequence consists of 1200 frames with a resolution of 720 × 560 pixels. We successfully detected objects with an inter-frame velocity as low as 1 pixel per frame. The false alarm rate, misdetection rate, and the accuracy rate of detection are 2%, 2.6% and 95.4%, respectively. The detection of each frame takes 38 ms, meeting the real-time requirement of object detection in videos with a frame rate of 25 frames per second.