Particulate pollution has become increasingly critical and threatening for human health. Although a number of approaches have been attempted for particulate pollution monitoring, these approaches are either expensive, unscalable, or requiring deployment of yet-another sensing infrastructure. In this study, by combining the advanced image dehazing and support vector machine techniques, we propose a novel particulate matter sensing approach using commercially off-the-shelf cameras. Using a Raspberry Pi-based testbed, we conducted a half-year measurement and conduct a comprehensive analysis of our approach. We show that our approach is effective, and the 80%-th estimation error is below 20 and 30 μg/m3 for PM2.5 and PM10 estimation, respectively. Moreover, the proposed approach can be easily applied to the existing camera surveillance infrastructure, as long as the photos contain both long-range and near-view objects.