Fast space debris detection is valuable to space safety and space vehicle operation. In this paper, we proposed a fast space debris detection method with grid-based learning. The image divided into 14×14 grids, and then the fast grid-based neural network (FGBNN) is used to locate the position of the space debris in the grids. The experiments show that, In the test set, the model can reach the accuracy of 98.83% and recall of 95.99% with SNR 2 of space debris, it can process 430 images per second with the given experiment settings in the prediction process. It is proved that the proposed method has excellent performance and fast detection speed.
Landslides are a natural geological phenomenon and often cause economic losses, property damage, and loss of lives. The detection and recognition based on the dense point cloud derived from Unmanned Aerial Vehicle (UAVs) images play an important role in understanding the mechanism of landslides. However, the brightness difference of UAV images with large altitude differences usually makes the dense point cloud within a landslide area show color stratification, and the processing results of popular commercial and open-source software are not ideal. In this study, a multi-view UAV image brightness compensation method based on regular 3D points is proposed, and is combined with the open-source software OpenMVG and OpenMVS to obtain a dense point cloud with balanced color on the landslide’s surface. The selected study area is located in the active landslide region caused by the Tengqing coal mining area in Guizhou Province, China. The terrain following flight method was used to acquire complete high resolution UAV images of the study area. Experimental results show that the global gain and offset compensation methods proposed in this study can effectively reduce the 22.3% brightness difference in the overlapping area of the adjacent images. The comparison results and the processing results of other commercial software show that the method can be applied to the color balance of the dense point cloud for UAV images with large-altitude differences.
The Gaofen-7 (GF-7) satellite is equipped with two area array sensor footprint cameras to capture the laser altimeter spot. In order to establish a direct correspondence between the laser data and the stereo image data, a new method is proposed to fit the center of the spot using the brightness difference between the spot image and the footprint image. First, the geometric registration between the spot image and the footprint image is completed based on feature matching or template matching. Then, the brightness values between the two images are extracted from the corresponding image position to form a measurement, and the least squares adjustment method is used to calculate the parameters of the brightness conversion model between the spot image and the footprint image. Finally, according to the registration relationship, the center of the identified spots is respectively positioned in the footprint images, so that the laser spots are accurately identified in the along-track stereo footprint images. The experimental results show that the spot error of this method is less than 0.7 pixel, which has higher reliability and stability, and can be used for a GF-7 satellite footprint camera.
1 January 2021 in Mathematical Biosciences and Engineering
Three-dimensional (3D) sparse reconstruction of landslide topography based on unmanned aerial vehicle (UAV) images has been widely used for landslide monitoring and geomorphological analysis. In order to solve the isolated island phenomenon caused by multi-scale image matching, which means that there is no connection between the images of different scales, we herein propose a method that selects UAV image pairs based on image retrieval. In this method, sparse reconstruction was obtained via the sequential structure-from-motion (SfM) pipeline. First, principal component analysis (PCA) was used to reduce high-dimensional features to low-dimensional features to improve the efficiency of retrieval vocabulary construction. Second, by calculating the query depth threshold and discarding the invalid image pairs, we improved the efficiency of image matching. Third, the connected network of the dataset was constructed based on the initial matching of image pairs. The lost multi-scale image pairs were identified and matched through the image query between the connection components, which further improved the integrity of image matching. Our experimental results show that, compared with the traditional image retrieval method, the efficiency of the proposed method is improved by 25.9%.