Engineering vehicle recognition based on video surveillance is one of the key technologies to assist illegal land use monitoring. At present, the engineering vehicle recognition mainly adopts the traditional deep learning model with a large number of floating-point operations. So, it cannot be achieved in edge devices with limited computing power and storage in real-time. In addition, some lightweight models have problems with inaccurate bounding box locating, low recognition rate, and unreasonable selection of positive training samples for the small object. To solve the problems, the paper proposes an improved lightweight Yolo-Fastest V2 for engineering vehicle recognition fusing location enhancement and adaptive label assignment. The location-enhanced Feature Pyramid Network (FPN) structure combines deep and shallow feature maps to accurately localize bounding boxes. The grouping k-means clustering strategy and adaptive label assignment algorithm select an appropriate anchor for each object based on its shape and Intersection over Union (IoU). The study was conducted on Raspberry Pi 4B 2018 using two datasets and different models. Experiments show that our method achieves the optimal combination in speed and accuracy. Specifically, the mAP50 is increased by 7.02 % with the speed of 11.24FPS under the engineering vehicle data obtained by video surveillance in a rural area of China.

Recent years have witnessed a growing interest in using WLAN fingerprint‐based methods for the indoor localization system because of their cost‐effectiveness and availability compared to other localization systems. In this system, the received signal strength (RSS) values are measured as the fingerprint from the access points (AP) at each reference point (RP) in the offline phase. However, signal strength variations across diverse devices become a major problem in this system, especially in the crowdsourcing‐based localization system. In this paper, the device diversity problem and the adverse effects caused by this problem are analyzed firstly. Then, the intrinsic relationship between different RSS values collected by different devices is mined by the linear regression (LR) algorithm. Based on the analysis, the LR algorithm is proposed to create a unique radio map in the offline phase and precisely estimate the user’s location in the online phase. After applying the LR algorithm in the crowdsourcing systems, the device diversity problem is solved effectively. Finally, we verify the LR algorithm using the theoretical study of the probability of error detection. Experimental results in a typical office building show that the proposed method results in a higher reliability and localization accuracy.

To accurately reconstruct the three-dimensional (3D) surface of dynamic objects, we proposed a wrapped phase extraction method for spatiotemporal analysis based on 3D wavelet transform (WT). Our proposed method uses a 2D spatial fringe image combined with the time dimension and forms a 3D image sequence. The encoded fringe image sequence’s wrapped phase information was extracted by 3D WT and complex Morlet wavelet, and we improved the wrapped phase extraction’s accuracy by using the characteristics of spatiotemporal analysis and a multi-scale analysis of 3D WT, then we reconstructed the measured object by wrapped phase unwrapping and phase height transformation. Our simulation experiment results show that our proposed method can further filter the noise in the time dimension, and its accuracy is better than that of the one- (1D) and two-dimensional (2D) WT wrapped phase extraction method and the 3D Fourier transform wrapped phase extraction method because the reconstructed spherical crown’s RMSE value does not exceed 0.25 and the PVE value is less than 0.95. Our results show that the proposed method can be applied to the dynamic 3D reconstruction of a real human thoracic and abdominal surface, which fluctuates slowly with respiration movement, further verifying its effectiveness.

Depth estimation of endoscopic images has always been a hot and difficult problem in minimally invasive surgery. Accurately estimating the depth information of endoscopic images can help surgeons improve the efficiency of surgical operations. This paper proposes a depth estimation method for endoscopic images based on deep learning, the ResNet34 module, FPN module and position information enhancement module are introduced to extract features. The reprojection loss, structural similarity index and edge-aware smoothing loss are used as loss function to improve the accuracy of depth prediction model and improve the applicability of the proposed method. The evaluation experiment uses the endoscopic image dataset, and the experimental results show that the depth information obtained by our method is more accurate than other state-of-the-art methods, which further verifies the effectiveness and advantages of the proposed method.