In the uncooled infrared imaging systems, owing to the non-uniformity of the amplifier in the readout circuit, the infrared image has obvious stripe noise, which greatly affects its quality. In this study, the generation mechanism of stripe noise is analyzed, and a new stripe correction algorithm based on wavelet analysis and gradient equalization is proposed, according to the single-direction distribution of the fixed image noise of infrared focal plane array. The raw infrared image is transformed by a wavelet transform, and the cumulative histogram of the vertical component is convolved by a Gaussian operator with a one-dimensional matrix, in order to achieve gradient equalization in the horizontal direction. In addition, the stripe noise is further separated from the edge texture by a guided filter. The algorithm is verified by simulating noised image and real infrared image, and the comparison experiment and qualitative and quantitative analysis with the current advanced algorithm show that the correction result of the algorithm in this paper is not only mild in visual effect, but also that the structural similarity (SSIM) and peak signal-to-noise ratio (PSNR) indexes can get the best result. It is shown that this algorithm can effectively remove stripe noise without losing details, and the correction performance of this method is better than the most advanced method.
Due to the limitation of less information in a single image, it is very difficult to generate a high-precision 3D model based on the image. There are some problems in the generation of 3D voxel models, e.g., the information loss at the upper level of a network. To solve these problems, we design a 3D model generation network based on multi-modal data constraints and multi-level feature fusion, named as 3DMGNet. Moreover, 3DMGNet is trained by self-supervised method to achieve 3D voxel model generation from an image. The image feature extraction network (2DNet) and 3D feature extraction network (3D auxiliary network) are used to extract the features of the image and 3D voxel model. Then, feature fusion is used to integrate the low-level features into the high-level features in the 3D auxiliary network. To extract more effective features, each layer of the feature map in feature extraction network is processed by an attention network. Finally, the extracted features generate 3D models by a 3D deconvolution network. The feature extraction of 3D model and the generation of voxelization play an auxiliary role in the training of the whole network for the 3D model generation based on an image. Additionally, a multi-view contour constraint method is proposed, to enhance the effect of the 3D model generation. In the experiment, the ShapeNet dataset is adapted to prove the effect of the 3DMGNet, which verifies the robust performance of the proposed method.
The uneven mass distribution of gyro rotors results in vibration, rotation and drift of gyro rotors, which seriously affect the performance index and life of gyro rotors. However, because there is no rigid connection between the rotary shaft and the shell of the gyro rotor, the dynamic balancing machine can only balance the vibration component of the single braced frame gyroscope rotor, and can’t measure the gyro rotor’s rotation component. By analyzing the influence of uneven rotor mass distribution on the gyro rotor performance, a method of eliminating two rotational degrees of freedom of the gimbal in gyro rotor by mandrel is proposed, which makes the dynamic balancing machine directly measure the vibration component and the moving component of the gyroscope rotor, and simplifies the dynamic balancing debugging process.
Camera calibration is one of the key steps of 3D reconstruction. Aiming at improving the accuracy and solving the complex problems of camera calibration, based on the camera model, a practical approach of camera calibration based on an isosceles right triangle planar pattern was presented. The camera coordinate system was set as the world coordinate system. Five camera intrinsic parameters needed to be calibrated. The planar pattern contained three apexes of the triangle and two arbitrary points on two right angle sides. More than six pattern images were captured from different poses, the improved Harris detector was used to obtain the sub-pixel coordinates of the corners. Established the constraint equations according to the right angle, camera intrinsic parameters can be calculated. This method reduced the requirements of pattern and was simple and convenient to operate. Experiment shows that the approach is accurate.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.