| Summary: | 碩士 === 國立臺北科技大學 === 能源與冷凍空調工程系 === 107 === Humans spend most of their time in homes, offices and other indoor environments. As a result, poor indoor air quality can cause physical discomfort to humans and even cause a variety of discomforts. Common sources of indoor air pollution may include smoking, cooking, or accumulation of carbon dioxide exhaled by organisms due to infrequent window opening. If human do not pay attention to indoor ventilation, it is easy to cause the amount of indoor carbon dioxide or carbon monoxide to exceed the standard value. If these two kinds of air pollutants exceed the criteria, it is easy to cause symptoms such as fatigue, headache, nausea, etc., and above symptoms may not be immediately detected by ordinary people due to air pollutants. This study is devoted to the image recognition of indoor planting and computer vision, and the trend chart is used to determine whether the carbon dioxide or carbon monoxide content in the indoor planting position has the same trend as the identification result. During the experiment, the author observed the changes of the plant at the four different concentration stages of the target gas, which were twice the standard value, the standard value, three times the standard value and the normal background value. In the carbon dioxide experiment, the data indicated that the RGB variation of the dominant colors is more correlated with the trend of carbon dioxide content.
|
|---|
