Signs of NY Chinatown

Observing Change of Signs of Chinatown in NYC

Genome of the Built Environment: Measuring the Unseen, Spring 2016, Harvard GSD
Teachers: Allen Sayegh, Stefano Andreani
Students: NJ Namju Lee, Keebaik Sim

Migration started from a long time ago and people move to other countries with unique culture from their origin. The physical representation of their culture creates sympathy among those immigrants who share the same culture and recognition of their voice in a new land.[1] Our goal is to analyze how immigrants maintain, lose their unique culture in a new environment. Visual cues are a key factor that represents one’s unique identity. Visual cues are well acquainted to people so that they not recognize visual cues mindfully. Chinatown in NYC is a perfect case study because of its long history and covers the largest area among different neighborhoods in NYC. Also, it has dominant cultural artifacts compared it’s neighbors, SoHo and little Italy.

To begin our study, we used openCV to compare the amount of Chinese Cultural elements are contained in the image of Chinatown by comparing it with the image of China. We found that this was a much larger task than intended, as the visual cues were such a broad concept so that it was difficult to conduct a research effectively. Instead, we decided to analyze the signs in Chinatown. To analyze the signs of Chinatown we used, OpenCV, an algorithm, which compare two images and detects common features and colors across each image has utilized. This algorithm has adopted to quantify the similarity and number of matching visual cues in each image. We collected hundreds of images of Chinatown by scraping Google Street View with Python and Open Street Map. These images ranged from the ground floor to four or five stories high, including areas that were usually unseen by human beings due to their visual limitations. Each of these collected images was then compared with another image set containing representational visual cues of China, previously assigned by us. However, there were several limitations. Firstly, OpenCV was very good at detecting outlines and patterns in two images, such as windows, doors, and leaves. It was not precise in identifying subtle elements. For instance, its ability to detect the similarities and differences between Chinese letter and alphabet was very poor, which is important for this research. Secondly, the subject, visual cues were too broad and needed to narrow down.

To overcome these limitations, we decided to focus our attention on the signs. They are one of the dominant visual queues that represent the identity of the neighborhood. Comparing to other visual queues such as doors, window and facades, signs are less regulated and can be easily changed depending on the characteristic of the store. Signs are composed of letter, color, material, and shape. Among these elements, letter and color strongly express the identity of culture. The proportions of the Chinese letters or alphabet can be use to identify if the sign signifies Chinese identity. There are five dominant colors that represent China such as black, white, yellow, blue, and red. To determine if these signs signified Chinese culture we analyzed the percentage of the sign covered by these unique colors.

Because Google Street View provides images from 2004 to 2014, we were able to find trends throughout time. The changes in letters and colors of the signs of the same site in 2007 and 2014 reveals another interesting findings. These data indicates the Chinatown in NYC is losing, maintaining, or accumulating their own unique identity. This could also possibly indicate how Chinatown’s signs will look like in the future. There are many factors that can explain the change of signs in an area with a strong cultural identity: gentrification, economic growth, the financial crisis in 2008, and so on. Analyzing the changes in signs could lead to other researches in the future.

[1] Rancière, Jacques. The Politics of Aesthetics : The Distribution of the Sensible. London ; New York: Continuum, 2004.

 

 

Font Resize