End-to-end(E2E) scene text recognition (the joint detection and recognition of natural text images) is developing rapidly, and the joint optimization strategy for image sequence alignment has become a research hotspot. Those existing methods are either difficult to train or costly for character annotations. In this paper, a novel end-to-end scene text recognition framework is proposed, Based on the Swin-Transformer (Swin-T) FPN backbone network, the model adopts the instance segmentation method to obtain the text mask and binarizes it to directly locate its polygon boundaries. Meanwhile, to solve the problem of low fitting efficiency of the text sequence recognition module, we designed a self-monitoring Mask-Supervised Attention (MSA) mechanism to accelerate the fitting speed and fitting accuracy of the recognition module, thereby improving the joint performance for E2E text recognition. The results show that in the E2E text recognition task, the F-measure performance of the proposed model achieves not only 2.3%, 2.7% and 11.9% improvement on ICDAR 2015 on strong, weak and generic lexicons, but also 4.8%, 9.9% improvement on Total-text on full and none lexicons compared with other typical models.
In this paper, a deep convolution neural network for image registration using homography transformation is proposed to improve the speed and accuracy of image registration. The four-corner homography parameterization is carried out by randomly clipping and perturbing the image block, and then the mapping from one image to another is completed to form the homography image registration dataset. In this network architecture, the homography matrix is obtained by returning the mean square error to the corner variables of the local region. In the preprocessing stage, the image is equalized by the histogram and the feature is magnified. The trained homography matrix is used for the affine transformation of the registered image to verify the effectiveness of the model. We test the dataset of homography image registration and experiment on various noises and various image enhancement effects. We also compare several traditional algorithms. The results prove that the accuracy of the proposed model is state-of-art. The processing speed of a single image is only 0.28 seconds, which has strong noise adaptability and the best performance.
As one of the most important transportation, the safety of railway is paid much attention to. The quality of wheel should
be checked periodically, especially in high-speed application. Normally, Non Destructive Testing (NDT), such as
ultrasonic inspection method, is applied on wheels to find the defect. A stationary automatic railway wheelset inspection
system by using ultrasonic technique is described in this paper. The phased array ultrasonic technique and wheel defect
inspection method is described in detail. Specially designed line is installed for wheelset transportation. Wheelset lifting
and rotating device is used for wheelset loading, unloading and rotating. A steel frame with complicated mechanical
structure and ultrasonic devices are designed for wheelset defect detecting. System ultrasonic performance, system
working flow, system control networking, data processing and results displaying are also described in the paper. Now,
the system is installed in Chinese EMU maintenance center for disassembled wheelset inspection and the safety of
wheels is well protected.
The railway line profile measurement is necessary for the safety of the train. This article expounds a method of railway line profile measuring using laser ranging and laser scanning technology with high precision and speed. With this method, the obstacle near the track can be found out and the hidden trouble can be removed. In tunnel, the crack and deformation on the tunnel wall can be measured. The parameter of the track and contact wire can be also inspected, such as rail gauge and superelevation, position of contact wire (stagger and height), wire wearing.
Flange Height and Flange Thickness are vital parameters for train wheels. Usually, they are manually measured by wheel
vernier after train stopped. This article introduces a dynamic wheel inspection system. Pulsed slit diode lasers project
onto wheel tread while train is in motion, CCD cameras record light-section images of wheel profile. Geometric
parameters and profile of wheel are measured on basis of image processing. This paper introduces measuring principle
and structure of this inspection system, analyzes influences resulted from laser parameters and environment, designs
instantaneous high-powered slit diode laser and narrowband optical filter according to ±0.2mm accuracy requirement
and environment situation.
In this paper, the importance of the contact line height and gradient in the electrified railway and the current inspection methods for the contact line height and gradient are analyzed, and then the dynamic detection system for that is deigned, which based on the laser phase ranging principle. The detection system is setting on the top of locomotive and a cooperative target is fixed on the pantograph; the laser system measures the height between the top of locomotive and the working pantograph by the cooperative target when the locomotive runs, and then the gradient of contact line can be calculated in real time when the locomotive's running information is provided. The laser phase ranging system uses the DFT method to calculate the phase difference, which can get the higher resolution than the method of the electronic phase demodulation and reduce the influence of the shift of laser intensity etc. The dynamic detection system works well to detect the contact line gradient, without influencing the normal operation of the locomotive, and the disadvantages of manual detecting and detecting car are avoided.
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.