Person reidentification (re-id) aims to match a specified person across non-overlapping cameras, which remains a very challenging problem. While previous methods mostly focus on feature extraction or metric learning, this paper makes the attempt in jointly learning both the global full-body and local body-parts features of the input persons with a multichannel convolutional neural network (CNN) model, which is trained by an adaptive triplet loss function that serves to minimize the distance between the same person and maximize the distance between different persons. The experimental results show that our approach achieves very promising results on the large-scale Market-1501 and DukeMTMC-reID datasets.
Soil mechanical properties play the most important role for the structural safety. But soil itself develops with environment as climate, loading and even surrounding biochemical contents which will strongly change the soil mechanical properties and may bring drastic damage to structural foundation. In-time monitoring and estimating on soil mechanical properties is proposed. Two piezoceramic transducers are embedded in predetermined locations: one is used as the actuator and the other is used as a sensor. The active-sensing method is applied to excite a stress wave propagating between the transducers. Variation of the wave velocity along the wave propagation path can be read while the soil properties changing. In this paper, soil specimens with different dry density, moisture content and soil granite ratio are tested to investigate the wave propagation variation through the soil with different properties. A model of shear wave velocity with different soil properties is established. Experimental results demonstrate that piezoelectric response wave may be potentially used to estimate soil physical properties.
In seasonally frozen soil regions, the frost heaving problem made it difficult to monitor or evaluate the pile safety for long term. So far, no mature tool can be utilized to monitor the frost heaving force, which was unevenly distributed along the pile. In this paper, a piezoceramic sensing based transient excitation response approach was proposed to monitor the frost heaving force in real time. Freeze-thaw cycles can result in great changes of soil engineering properties, including the frost heaving force. So, the freeze-thaw cycle was repeated fourth to study its effect. In the experiment, transient horizontal shock on the top of the pile will be detected by the 6 PZT sensors glued on the pile. The signal data received by the 6 PZT sensors can be used to illustrate the frost heaving force distribution along the pile. Moisture content effect is also one of the important reasons that cause the variation of soil mechanical properties. So three different moisture content (6%, 12%, 18%) testing soil were used in this experiment to detect the variance of the frost heaving force. An energy indicator was developed to quantitatively evaluate the frost heaving force applied on the pile. The experimental results showed that the proposed method was effective in monitoring the uneven distribution of frost heaving force along the pile.
In this paper, a hybrid approach is proposed to detect texts in natural scenes. It is performed by the following steps:
Firstly, the edge map and the text saliency region are obtained. Secondly, the text candidate regions are detected by
connected components (CC) based method and are identified by an off-line trained HOG classifier. And then, the
remaining CCs are grouped into text lines with some heuristic strategies to make up for the false negatives. Finally, the
text lines are broken into separate words. The performance of the proposed approach is evaluated on the location
detection database of ICDAR 2003 robust reading competition. Experimental results demonstrate the validity of our
approach and are competitive with other state-of-the-art algorithms.
Dim small target detection, which is characterized by complex background and low Signal-to-Noise Ratio (SNR), is critical
for many applications. Traditional detection algorithms assume that noise is not useful for detecting targets and try to
remove the noise to improve SNR of images using various filtering techniques. In this paper, we introduce a detection
algorithm based on Stochastic Resonance (SR) where stochastic resonance is used to enhance the dim small targets. Our
intuition is that SR can achieve the target enhancement in the presence of noise. Adaptive Least Mean Square (ALMS)
filtering is first adopted to estimate the background, and the clutter is suppressed by subtracting the estimated background
image from the source image. Adaptive SR (ASR) method is then employed to enhance the target and improve the SNR
of the image containing the target and noise. ASR tunes and adds the optimal noise intensity to increase the power of the
targets and therefore improve the SNR of the image. Several experiments on synthetic and natural images are conducted to
evaluate our proposed algorithm. The results demonstrate the effectiveness of our algorithm.
Concrete piles are widely used in the construction of civil infrastructures and it is important to perform the health
monitoring of concrete piles for safety purposes. In this paper, a piezoceramic-based innovative approach is proposed for
the damage detection and health monitoring of concrete piles. A multi-functional piezoceramic-based transducer device,
the smart aggregate, is developed for health monitoring purposes. An active-sensing network is formed by embedding
the proposed smart aggregates at the pre-determined locations in the concrete piles before casting. In the proposed
approach, one smart aggregate is used as an actuator to excite the desired waves and the other distributed smart
aggregates are used as sensors to detect the wave responses. An energy distribution vector is formed based on the
wavelet-packet analysis results of sensor signals. A damage index is formed by comparing the difference between the
energy distribution vectors of the health concrete pile and that of the damaged concrete pile. To verify the effectiveness
of the proposed approach, two concrete piles instrumented with smart aggregates are used as testing objects. One
concrete pile is intact and the other has a man-made crack in the middle of the pile. Experimental results show that the
there are differences between the energy distribution vectors of the damaged pile and that of the intact pile due to the existence of the crack. The proposed method has the potential to be applied to perform automated integrity inspections for new piles and for the long-term health monitoring of piles in services.
KEYWORDS: Bridges, Sensors, Inspection, Ferroelectric polymers, Analog electronics, Data processing, Analytical research, Finite element methods, Data acquisition, Software development
Economic and effective evaluation of the actual condition of a bridge is an important issue for bridge maintenance and preservation. The current condition of the bridge is usually far more different from the construction design and also with the bridge being in service for few years, a quantitative evaluation is essential. Assimilating reviews and comparisons of traditional inspection and evaluation methods, this paper promotes the concept of application of wireless sensors developed at the Bridge Research Center for rapid installation and low power requirements for in-service real-time bridge inspection. The concept focuses on obtaining useful data such as strain, displacement, frequency for estimating the actual characteristics of the bridge while in service, thus avoid closing the traffic as compared to traditional load test which needs traffic closure causing inconvenience to public and indirectly affecting economy. Plans and procedures of the field inspection are detailed as well as the data processing and the analysis results are presented. The effectiveness and feasibility of the proposed real-time inspection based on wireless sensors approach are illustrated via the practical inspection of a deck beam bridge.
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.