Morpho-Colorimetric Characterization of the Sardinian Endemic Taxa of the Genus Anchusa L. by Seed Image Analysis

In this work, the seed morpho-colorimetric differentiation of the Sardinian endemic species of Anchusa (Boraginaceae) was evaluated. In Sardinia, the Anchusa genus includes the following seven taxa: A. capellii, A. crispa ssp. crispa, A. crispa ssp. maritima, A. formosa, A. littorea, A. montelinasana, and A. sardoa. Seed images were acquired using a flatbed scanner and analyzed using the free software package ImageJ. A total of 74 seed morpho-colorimetric features of 2692 seed lots of seven taxa of Anchusa belonging to 17 populations were extrapolated and used to build a database of seed size, shape, and color features. The data were statistically elaborated by the stepwise linear discriminant analysis (LDA) to compare and discriminate each accession and taxon. In addition, the seed morpho-colorimetric differences among coastal and mountainous taxa were evaluated. Considering the ecological conditions, the LDA was able to discriminate among the Anchusa taxa with a correct identification of 87.4% and 90.8% of specimens for mountainous and coastal plants, respectively. Moreover, the LDA of the 17 populations of Anchusa showed a low separation among species and populations within the coastal group, highlighting how the long-distance dispersal by flotation on the sea water surface and the pollination network may influence the similarity patterns observed. In addition, a misattribution was observed for A. crispa ssp. crispa, which was misclassified as A. crispa ssp. maritima in 14.1% of cases, while A. crispa ssp. maritima was misidentified as A. crispa ssp. crispa in 21.1% of cases, highlighting a close phenotypic relationship between these two taxa. The statistical results obtained through the seed image analysis showed that the morpho-colorimetric features of the seeds provide important information about the adaptation and evolution of Anchusa taxa in Sardinia.

Pixel-wise Colorimetric Characterization based on U-Net Convolutional Network

In this article, we present a U-Net convolutional network for solving insufficient data problems of color patches in colorimetric characterization. The U-Net network uses data augmentation annotated over 6,885,222 colors, 32,027,200 color patches, and 2,098 billion pixels directly from only eight standard colorimetric images of ISO 12640 (CIELAB/SCID). By applying the U-Net network trained on big augmented data, the pixel-wise colorimetric characterization is implemented from digitalized red, green, blue image samples to ISO 12640 (CIELAB/SCID) CIELAB standard colorimetric images. The performance efficiency of the U-Net network is superior to that of the convolutional neural network on both training and validating epochs. Moreover, pixel-wise color colorimetric characterization is achieved using the intelligent machine vision of U-Net integrated with a data augmentation technique to overcome the drawback of complex color patches and labor-intensive tasks. This study might improve colorimetric characterization technology with a resolution of 2560-by-2048 for over 4 million pixels. The results reveal that U-net with pixel-wise regression enhances the precise colors of images, taking detail and realism to a new level.

A Comparison of Colorimetric Performance of Oculus and HTC Virtual Reality Headsets

The colorimetric characterization of the two virtual reality headsets, namely, HTC and Oculus are compared to each other. In order to do that, first, a colorimeter is used to measure the colorimetric values of the primary ramps in a darkened and controlled environment. It is observed that the two headsets behave more or less the same with HTC outputting an overall higher level of luminance and having more consistent right and left displays. Weighted regression is also used as a means to characterize the devices and the results are compared to the traditional method of colorimetric characterization showing the superiority of the weighted regression in this case.