In this experiment, I was trying out the concept from a previous exercise where RGB values are being measured and determined with a threshold that was set. I also tried color coding the dominant color pixel of the object to being normal,suspicious and dangerous as it was used for existing applications.The intention of the experiment was to develope the idea of machine biasness and behind these machine learning lies with how data are used to generate informations and outputs.
Program used: ChatGPT, Python, OpenCV
Using OpenCV as the program for computer vision, it's few main usage is for object detection and video process. Using the functions available, I incorporated the code with colored-pixel detection and with a threshold limit.
The narrative intended was to create a data value scanner using computer vision that evaluates through the amount of dominant color pixels in an image with the “what if” scenario of selling your daily data of anything as commodity.
As supposed to a bigger project to communicate the idea of data as commmodity and data privacy, it was simple experiment influenced by the works of "Data Production Labor". However through this experiment, it was getting to know the capabilities of computer vison and how possibly the data are used to create for applications. The idea of training datasets from collected personal data can prove to be biased base on who are the ones setting the limitations and "rules" to trigger events.