02/ COLORS OF AR FILTERS
METHOD AND ANALYSIS
RESEARCH QUESTION
The main purpose of the research was to determine the key role of color in filters. Specifically, to identify which are the main trends in the use of colors between the two aesthetics and to map when it’s actually used as a tool of user’s face alteration
DATA COLLECTION
The analyzed images have been taken from a sample of the same dataset of the first research (analyzing the type of modifications among filters). From a total of 1.240 images of @Instmask @face.effect @beauty.filters @aestheticfilters, 940 posts have been taken in analysis. Specifically, all images were divided into two main macrocategories, according to the results of the research one:
the first macrocategory contains all perfectionist beauty posts (derived from @beauty.filters and @aesthetic.filters profiles, 470 images) while the second macrocategory contains all dystopyc beauty posts (derived from @instmask and @face.effect profiles, 470 images).
ANALYSIS
1.Image sorter
The first part of the analysis was done by loading the images in the Image sorter tool which clusters them by color similarity, creating a grid composition of images. The process was made separately between the two macrocategories realizing two different grid of images in order to be compared.
1.Color summarizer
The second part of the analysis aimed to identify the main color code of each image. The analysis of each post was done using the Color summarizer tool that produces descriptive color statistics. The image was divided into 3 main color clusters. The one with the highest percentage of pixels of the same color was saved and integrated in the composition of images created before.
The final part of the analysis was the comparison between the two color grids to identify the main differences between the aesthetics. In addiction, all posts having as their main color code a special one
different from natural skin, has been analyzed mapping when the special color was affecting the face or the background.
