The atmospheric quality monitor is an essential component of atmospheric protection. However, using complicated large-scale chemical equipment required by traditional methods is difficult for ordinary people. We are committed to allowing ordinary people to obtain a local air quality index (AQI) and then participate in the air quality monitoring of the entire district quickly and easily. So, we propose a method improved from Transformer in this article. This method processes pictures taken by mobile devices and predicts a more accurate local AQI level. The proposed method performs better than traditional methods in frequency and flexibility. Tokens in the Double Output Vision Transformer (DOViT) proposed in this article will be processed with the multihead self-attention (MSA) mechanism of the model to automatically extract the features in pictures for achieving higher classification accuracy. The application of Transformer makes our model more concise and efficient than previous CNN-based methods.

Air pollution detection and measurement is an important problem. Fast and effective explanation of the air quality is a necessary technique for monitoring the air pollution. However, the existing air quality measurement devices severely rely on many sensors, which are not only expensive but also inconvenient to carry. With the development of deep learning technology, computer vision-based task, such as the video processing and understanding have achieved great progress. Recently, image-based air quality measuring methods have been proposed and achieved satisfying accuracy in specific scenes. Whereas, the performance of those methods are not stable due to the noises in images and missed temporal relations between single images. With the rising of the short-video platform, the acquisition and dissemination of video data becomes more convenient. To address the problems of image-based air pollution measurement, we propose a video-based dual-channel 3-D convolution network for stable and accurate measuring. Besides the basic visual channel, we add a semantic channel to guide the network learn region-level features. The features from two channels are combined for stable prediction. Moreover, to evaluate the effectiveness of our method, we collect an outdoor video data set for air quality measurement through smart phones. Extensive experimental results show that our method achieves good results in the air quality measurement and outperforms the image-based methods.

Recently, air quality analysis based on image sensing devices has attracted much attention. Since most smoke images in real scenes have challenging variances, which is difficult for existing object detection methods. To keep real-time factory smoke under efficient and universal social supervision, this paper proposes a mobile-platform-running efficient smoke detection algorithm based on image analysis techniques. We introduce the two-stage smoke detection (TSSD) algorithm based on the lightweight detection framework, in which the prior knowledge and contextual information are modeled into the relation-guided module to reduce the smoke search space, which can therefore significantly improve the performance of the single-stage method. Experimental results show that the proposed TSSD algorithm can robustly improve the detection accuracy of the single-stage method and the model has good compatibility for image resolution inputs. Compared with various state-of-the-art detection methods, the accuracy $$AP_{mean}$$A P mean of our proposed TSSD model reaches 59.24 $$\%$$% , even surpassing the current detection model Faster RCNN. In addition, the detection speed of our proposed model can reach 50 ms (20 FPS), meeting the real-time requirements. This knowledge-based system has the advantages of high stability, high accuracy, fast detection speed. It can be widely used in some scenes with smoke detection requirements, such as on the mobile terminal carrier, providing great potential for practical environmental applications.

Air quality monitor is important for both the environment and human health. In recent years, many computer vision-based air quality monitoring methods have been proposed and have achieved good accuracy in specific scenarios. However, current deep-learning-based models' deployment on mobile devices is impeded due to their excessively large size. To build a lightweight model, we investigate the potential of dilated convolutional kernels, which can effectively increase the receptive field and acquire multiscale information while maintaining model size. Eventually, we propose a novel method termed multiscale dilated network. The network is tailored to the characteristics of the air quality dataset by rearranging the dilate rate of the dilated convolutional kernels to focus on the global and semantically higher level features, thus improving the detection efficiency and accuracy of the model. Our proposed multiscale dilated block also includes a multiscale fusion mechanism that enables the model to extract air quality information at multiple scales simultaneously. Compared to the current state-of-the-art models, the multiscale dilated network integrates three different scales of dilated kernels, which leads to a reduction of 86.7% of parameters and 88.5% of floating point operations and achieves a top-1 classification accuracy of 94.2%. With the proposed model enhancements, we envision a future where mobile phones and IoT edge devices can function as mobile air quality monitoring stations, thereby creating a dynamic and distributed air quality map for air quality protection.