Object Detection in Optical Remote Sensing Images Based on FFC-SSD Model
Xue Junda, Zhu Jiajia, Zhang Jing, Li Xiaohui, Dou Shuai, Mi Lin, Li Ziyang, Yuan Xinfang, and Li Chuanrong
For the applications of efficient high-precision object detection in optical remote sensing (RS) images, this paper focuses on the difficulty of improving the detection accuracy of the SSD (single shot multibox detector) model on small and densely distributed objects in such images. An improved model FFC-SSD (multi-scale feature fusion & clustering SSD) is thereby proposed. For this purpose, a bounding-box group clustering (BGC) module is designed. Group clustering is implemented to obtain default object frame parameters that are more consistent with the size distribution of object samples and gives more attention to small objects. This module effectively improves the network’s ability to extract object locations. Then, an efficient de-pooling multi-scale feature fusion (MSFF) module is designed to enhance the ability of the model to extract object features and effectively reduce the efficiency loss of the model at the same time. The experimental results demonstrate the effectiveness and applicability of the FFC-SSD model for object detection in optical remote sensing images. The proposed model achieves a favorable balance between precision and efficiency and has high detection accuracy on small objects.
  • Jun. 20, 2022
  • Acta Optica Sinica
  • Vol. 42 Issue 12 1210002 (2022)
  • DOI:10.3788/AOS202242.1210002
Multi-image optical encryption method of JTC system combining CGH and frequency shift
Zheng Wei, Xi Sixing, Wang Guilin, Li Yonghong, and Jiang Qichang
In order to improve the practicability of JTC optical image encryption system, solve its noise problem, improve its encryption efficiency and security, a multi-image optical encryption method based on computer generated hologram (CGH) and the frequency shift of Fourier transform was proposed. Firstly, each image with different size and type was modulated by random phase and Fourier transform, then the frequency spectrum of multiple images was modulated by frequency-shift phase and superimposed to be encoded into binary real-value CGH, finally the CGH was encrypted into joint power spectrum by JTC optical image encryption system. In the decryption process, the encrypted image was decrypted by 4F system to obtain the CGH. The binary real-value CGH has strong noise resistance to eliminate the effects of noise. Then, multiple decrypted images can be obtained after Fourier transform. Simulation results show that the proposed method can encrypt and decrypt multiple images of different sizes and types in parallel, and has high encryption efficiency. Meanwhile, multiple images with mutual keys and double optical keys ensure the security of the encryption system.
  • Jun. 14, 2022
  • Infrared and Laser Engineering
  • Vol. 51 Issue 5 20220175 (2022)
  • DOI:10.3788/IRLA20220175
Laser altimetry data processing and combined surveying application of ZY3-03 satellite
Li Guoyuan, Tang Xinming, Zhou Ping, Chen Jiyi, Liu Zhao, Dou Xianhui, Zhou Xiaoqing, and Wang Xia
ZY3-03 is a land remote sensing satellite for 1: 50 000 stereo mapping built by the Ministry of Natural Resources. It is equipped with operational laser altimeter, which is mainly used to obtain high-precision elevation control points. In this paper, aiming at the laser altimetry data of ZY3-03 satellite, the standardized surveying processing flow and the method of extracting elevation control points is studied. Moreover, the accuracy verification in Sunid Right Banner of Inner Mongolia and Suzhou of Jiangsu Province is implemented, and the combined surveying and mapping application in Heilongjiang and Hebei areas is experimented and validated. The accuracy verification results show that the elevation accuracy of ZY3-03 laser points in the flat area of Sunid Right Banner in Inner Mongolia is (0.051±0.232) m, and the overall accuracy of the laser points in the urban area of Suzhou, Jiangsu is (0.414±6.213) m, and the elevation accuracy after elevation control points extraction is (-0.526±0.624) m, which can meet the elevation control requirement of 1:50 000 mapping; The application of combined surveying and mapping shows that the elevation accuracy of stereo images can be improved from 5.27 m to 2.58 m in flat area of Heilongjiang and from 11.25 m to 4.45 m in Taihang mountain area of Hebei by using laser elevation control points derived from the ZY3-03 satellite. It is concluded that the elevation accuracy of stereo images can be effectively improved by using laser elevation control points of ZY3-03 satellite in both flat and mountainous areas, and the requirement of 1:50 000 mapping can be met.
  • Jun. 14, 2022
  • Infrared and Laser Engineering
  • Vol. 51 Issue 5 20210356 (2022)
  • DOI:10.3788/IRLA20210356
Local neighborhood feature point extraction and matching for point cloud alignment
Wang Mingjun, Yi Fang, Li Le, and Huang Chaojun
Point cloud registration is one of the key technologies for 3D reconstruction. To address the problems of the iterative closest point algorithm (ICP) in point cloud matching, which requires high initial position and low speed, a point cloud registration method based on adaptive local neighborhood feature point extraction and matching was proposed. Firstly, according to the relationship between the local surface change factor and the average change factor, feature points were adaptively extracted. Then, the fast point feature histogram (FPFH) was used to comprehensively describe the local information of each feature point, the coarse alignment was achieved combining with the random sampling consistency (RANSAC) algorithm. Finally, according to the obtained initial transformation and feature point based ICP algorithm, the fine alignment was achieved. The alignment experiments were conducted on the Stanford dataset, noisy point cloud and scene point cloud. The experimental results demonstrate that the proposed feature point extraction algorithm can effectively extract the features of the point cloud, and by comparing with other feature point detection methods, the proposed method has higher alignment accuracy and alignment speed in coarse alignment with better noise immunity; compared with the ICP algorithm, the registration speed of the feature point based-ICP algorithm in the Stanford data set and scene point cloud is increased by about 10 times. In the noisy point cloud, the registration can be performed efficiently according to the extracted feature points. This research has certain guiding significance for improving the efficiency of target matching in 3D reconstruction and target recognition.
  • Jun. 14, 2022
  • Infrared and Laser Engineering
  • Vol. 51 Issue 5 20210342 (2022)
  • DOI:10.3788/IRLA20210342
Study on Vector Radiative Transmission Characteristics of Polarization Optics in Underwater Bubble Environment
Song Qiang, Sun Xiaobing, Liu Xiao, Ti Rufang, and Huang Honglian
Using optical imaging equipment for survey of underwater resources often obtains lower quality of image because of underwater bubble’s interference. Taking analysis of radiation transmission characteristics of bubble and interference characteristics on the optical imaging instrument in the water environment has important guiding significance for improving anti-jamming capability of underwater imaging. At first, the single bubble underwater imaging environment is built and the changing trends of radiation intensity and polarization state for light transmitting in single bubble are simulated, and the transmission phase function on the interface of single bubble is obtained. Then, the changing trends of radiation intensity and polarization state of light with bubble distribution under multi-bubble imaging environment are simulated when increasing the number of bubbles on the basis of research on single bubble. Finally, the effects on forward transmission and backward transmission characteristics with different radius of bubble sizes and different distances of light transmission are simulated with Monte Carlo method based on theory of geometrical optics approximation. The study of bubble theory, simulation and experiment shows that the changing trend of produced light’s polarization at surface of bubble has strong correlation with observation geometry, when the angle of incidence is small, light has strong penetrability and the attenuation of radiation intensity is slow. The degrees of polarization of forward transmission and backward transmission show significantly opposite trend with increase of transmission distance. The theory, simulation and experimental results are basically consistent. The research can provide a certain reference value in target polarization imaging field under complex underwater environment.
  • Jun. 07, 2022
  • Acta Optica Sinica
  • Vol. 42 Issue 12 1210001 (2022)
  • DOI:10.3788/AOS202242.1210001
Point Cloud Analysis Combining Gated Self-Calibration Mechanism and Graphical Convolutional Network
Xu Jiali, Fang Zhijun, and Wu Shiqian
Point clouds, unlike images represented by dense grids, are characterized by irregularity and disorder, making it difficult to precisely reason out the shape features in point cloud data. The internal-external shape son volution for point sets (IE-Conv) is proposed to address the limitations of current research. The local shape inside the point set is treated separately from the global shape outside the point set using an efficient bilateral design. Rich inter-point relationships are selectively studied in a gate-based manner within the point set, while point-by-point and local features are optimized by self-calibration functions; outside the point set, global shapes are constructed using graph convolution and focus on long-range dependencies between point sets. Finally, the organic fusion of the bilateral outputs is performed. This paper performs classification and segmentation experiments on the standard ModelNet40 and ShapeNet datasets by hierarchically embedding IE-Conv into the shape-reasoning convolutional network (SR-Net). The experimental results show that the classification task achieves an accuracy of 93.9% and the segmentation task achieves the mean intersection over the union of 86.4%, which verifies the good performance of SR-Net in point cloud analysis.
  • May. 23, 2022
  • Laser & Optoelectronics Progress
  • Vol. 59 Issue 12 1210017 (2022)
  • DOI:10.3788/LOP202259.1210017
Fine-Grained Image Recognition of Wild Mushroom Based on Multiscale Feature Guide
Zhang Zhigang, Yu Pengfei, Li Haiyan, and Li Hongsong
Deep learning technology is proposed to solve the social problem of the frequent occurrences of wild mushroom poisoning in China. However, due to the small difference between classes and complex image backgrounds, fine-grained recognition accuracy is low. To solve this problem, this paper proposes an improved ResNeXt50 network. First, a multiscale feature guide (MSFG) module is designed, which guides the network to learn and use low and high-level features fully through short connections. Then, the improved attention mechanism module is used to reduce the network’s learning for complex backgrounds. Finally, the different hierarchical features in the model are fused, and the obtained joint features are used for recognition. Experimental results show that the accuracy of the proposed network on the test set can reach 96.47%, which is 2.64 percentage points higher than the unimproved ResNeXt50 network. Comparison results show that the accuracy of the improved network model is 8.10 percentage points, 5.13 percentage points, 3.24 percentage points, 3.30 percentage points, and 4.25 percentage points better than VGG19, DenseNet121, Inception_v3, ResNet50, and ShuffleNet_v2, respectively.
  • May. 23, 2022
  • Laser & Optoelectronics Progress
  • Vol. 59 Issue 12 1210016 (2022)
  • DOI:10.3788/LOP202259.1210016
Remote Sensing Image Segmentation Using Super-Pixel and Dot Product Representation of Graphs
Zhang Daming, Zhang Xueyong, Liu Huayong, and Li Lu
The image segmentation method using division-combination mitigates the limitations of the traditional pixel-based remote sensing image segmentation algorithm, such as noise interference, low segmentation efficiency, and poor segmentation effect. Thus, this paper proposes a new split-merge-based remote sensing image segmentation method using the super-pixel and dot product representation of graphs. First, the image is divided into super-pixels using the simple linear iterative clustering (SLIC) algorithm. Second, the texture feature of each super-pixel area is measured and distance between any two areas is calculated with respect to spatial proximity. Third, each super-pixel area is mapped as a vertex of the graph. Therefore, the dot product representation of graphs is modified and used to construct a similarity matrix; thereafter, all vertices (i.e., super-pixel areas) are mapped as new vectors clustered by angular-based k-means algorithm to get the final segmentation results. The experimental results show that the proposed method has stable segmentation results, improves the accuracy of the segmentation, and achieves a better visual segmentation effect.
  • May. 23, 2022
  • Laser & Optoelectronics Progress
  • Vol. 59 Issue 12 1210015 (2022)
  • DOI:10.3788/LOP202259.1210015
Reconstruction of Magnetic Resonance Images Based on Dual-Domain Crossed Codec Network
Zhang Dengqiang, Liu Xiaohan, and Pang Yanwei
Magnetic resonance imaging (MRI) has outstanding soft-tissue contrast and provides unparalleled benefits in various diagnoses. It is an important way of observation in current clinical practice. The scanning period of an MRI, however, is long, which greatly limits the diagnostic efficiency. Obtaining undersampled K-space data through partial scanning at a specific acceleration magnification is a critical approach to save scanning time. Existing approaches only rebuild the K-domain or the image domain alone or alternately process the two domains through serially coupled image domain and K-domain convolution, resulting in poor reconstruction performance. A dual-domain parallel codec structure that processes image domain and K-domain data simultaneously is presented to provide high-quality reconstruction of undersampled K-space data at high acceleration rates. The proposed technique reconstructs the undersampled image domain and K-domain data using two parallel codec networks, respectively, then combines the features of the K-domain branch into the image domain using the inverse Fourier transform, considerably enhancing reconstruction quality. For presampling data with varying acceleration magnifications, experimental results reveal that the proposed method outperforms other U-Net-based image reconstruction methods. This proposed method is projected to develop into a high-performance, high-acceleration-magnification MRI undersampling data reconstruction method that can be used in clinical MRI reconstruction.
  • May. 23, 2022
  • Laser & Optoelectronics Progress
  • Vol. 59 Issue 12 1210014 (2022)
  • DOI:10.3788/LOP202259.1210014
Visual Tracking Combining Attention and Feature Fusion Network Modulation
Xu Keying, Shu Ping, and Bao Hua
The existing tracking algorithms for network modulation ignore high order feature information, so they are prone to drift when dealing with large scale changes and object deformations. An object tracking algorithm that combines the attention mechanism and feature fusion network modulation is proposed. First, an efficient selective kernel attention module is embedded in the feature extraction backbone network, so that the network pays more attention to the extraction of target feature information; second, a multiscale interactive network is used for the extracted features to fully mine the multiscale information in the layer, and high order feature information is fused to improve the ability of target representation, to adapt to the complex and changeable environment in the tracking process; finally, the pyramid modulation network is used to guide the test branch to learn the optimal intersection over union prediction to achieve an accurate estimation of the targets. Experimental results show that the proposed algorithm achieves more competitive results than other algorithms in tracking accuracy and success rate on VOT2018, OTB100, GOT10k, TrackingNet, and LaSOT visual tracking benchmarks.
  • May. 23, 2022
  • Laser & Optoelectronics Progress
  • Vol. 59 Issue 12 1210013 (2022)
  • DOI:10.3788/LOP202259.1210013