Visualization of gaseous chemical substances, such as odors, is very challenging, and few studies are being conducted to address this issue. These odor visualization technologies are expected to be more beneficial for various applications, such as detecting hazardous or human relating gas source localization. Currently, semiconductor gas sensors are commonly used for gas measurements; however, these sensors have disadvantages, such as slow response time and inability to detect multiple types of gases. Furthermore, it is difficult to understand how gases are distributed in space. Thus, we have developed a gas sensor utilizing localized surface plasmon resonance (LSPR). LSPR gas sensors enable real-time, high-speed response and recovery, and by analyzing wavelength information using a hyperspectral camera, gas imaging and identification become possible. In this letter, we conducted experiments on visualizing the spatiotemporal distribution of multiple gases and identifying gas space through the analysis of spectral information using a 2-D LSPR gas sensor and a hyperspectral camera. As a result, we successfully visualized and identified the distribution of each gas in a situation where multiple gases were present by performing principal component analysis analysis on the spectral data obtained from hyperspectral images.