Three-dimensional (3D) object detection is crucial for accurate recognition of autonomous driving roads, and the distribution of point clouds in 3D scenes becomes sparse with increasing distance, thus seriously affecting the sensor’s perception precision. To address this problem, we propose a two-stage 3D object detection network based on point and voxel feature fusion. In the first stage, a spatial semantic feature fusion module is designed to effectively fuse low-level spatial features and high-level semantic features to generate high-quality proposals. Then, an attention mechanism-based residual module is constructed to expand the receptive field and adaptively aggregate the voxel features in the 3D scene. At the same time, the sampled key points and voxel features are fused to extract the key information in the 3D scene. In the second stage, the graph network pooling module is introduced to construct local graphs on 3D proposals using key point features as nodes to estimate the confidence and location of objects more accurately. Experimental results on the KITTI dataset show that the detection precision is improved significantly in easy, moderate, and hard tasks.
Aiming at the subtle differences between fabric defects and complex texture background and the problem of small size defects, we propose a fabric surface defect recognition method based on multibranch residual network. First, the convolution kernel is decomposed and the multibranch shared feature extraction module is constructed using the residual connection to mine the feature information at different scales. At the same time, squeeze-and-excitation network (SE-Net) with a reduction ratio of 16 is added to the module to suppress the interference of complex texture background. Second, the channel attention mechanism is embedded in the residual block of the network backbone to realize the compensation for the lost feature information. Finally, the swish activation function is used to enhance the accuracy and robustness of the deep network, and transfer learning is used to further improve the network accuracy. The experimental results show that the proposed model is superior to the existing model in terms of average recognition accuracy. The proposed model has the highest recognition accuracy of 96.17% for eight defect categories, which proves the effectiveness of fabric surface defect class recognition.
The relationship between the transmission distance and the number of Fresnel zone plate rings is evaluated. The influence of different transmission distance and the number of Fresnel zone-plate rings are analyzed. Their parameters of the best reconstructed image are obtained. A point light source is taken as an example, a phase-only spatial light modulator with the resolution of 3840×2160 pixels and the pixel interval of 3.74 μm is used for experimental verification. The numerical simulation and optical experiment results show that the optimal reconstructed image has a transmission distance of 200 mm. Meanwhile, the optimal Fresnel zone ring number is 8 under the same conditions. This study provides the optimal parameters for the spatial light modulators with different size.
KEYWORDS: Digital holography, Holography, Tomography, Compressed sensing, Reconstruction algorithms, Holograms, Charge-coupled devices, Scattering, Signal to noise ratio, 3D modeling
Digital holography can reconstruct 3-D data cube from a 2-D hologram for the tomographic imaging. Digital one-shot inline holography (DOIH) maintains the maximal space-bandwidth product compared with off-axis holography and keeps both amplitude and phase in the interference pattern. DOIH often suffers from intrinsic defects such as twin-image interruption and squared noise. In this work, compressive sensing is applied in the tomographic reconstruction to overcome the defects. The designed algorithm based on compressive DOIH demonstrates the feasibility in removing the squared noise from a single 2-D in-line hologram.
Biochips have been an advanced technology for biomedical applications since the end of the 20th century. Optical detection systems have been a very important tool in biochip analysis. Microscopes are often inadequate for high resolution and big view-area detection of microarray chips, thus some new optical instruments are required. In this work, a novel digital imaging scanning system with dark-field irradiation is developed for some biomedical applications for microarray chips, characterized by analyzing genes and proteins of clinical samples with high specific, parallel, and nanoliter samples. The novel optical system has a high numerical aperture (NA=0.72), a long working distance (wd>3.0 mm), an excellent contrast and signal-to-noise ratio, a high resolving power close to 3 µm, and an efficiency of collected fluorescence more than two-fold better than that of other commercial confocal biochip scanners. An edge overlap algorithm is proposed for the image restructure of free area detection and correcting scanning position errors to a precision of 1 pixel. A novel algorithm is explored for recognizing the target from the scanning images conveniently, removing noise, and producing the signal matrix of biochip analysis. The digital imaging scanning system is equally as good for the detection of enclosed biochips as it is for the detection of biological samples on a slide surface covered with a glass cover slip or in culture solution. The clinical bacteria identification and serum antibody detection of biochips are described.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.