Research Article

Vision-Based Deep Q-Learning Network Models to Predict Particulate Matter Concentration Levels Using Temporal Digital Image Data

Table 3

Thumbnails of target regions where particulate matters are analyzed to construct the predictive model.

Category Thumbnails# of video sequences

Indoor office2,200

Outdoor parking lot of Konkuk Univ.3,200

Outdoor parking lot of Keimyung Univ.3,000

Mobile video clips1,000

Experimental chamber500