At present, most compressed sensing (CS) algorithms have poor converging speed, thus are difficult to run on PC. To deal with this issue, we use a parallel GPU, to implement a broadly used compressed sensing algorithm, the Linear Bregman algorithm. Linear iterative Bregman algorithm is a reconstruction algorithm proposed by Osher and Cai. Compared with other CS reconstruction algorithms, the linear Bregman algorithm only involves the vector and matrix multiplication and thresholding operation, and is simpler and more efficient for programming. We use C as a development language and adopt CUDA (Compute Unified Device Architecture) as parallel computing architectures. In this paper, we compared the parallel Bregman algorithm with traditional CPU realized Bregaman algorithm. In addition, we also compared the parallel Bregman algorithm with other CS reconstruction algorithms, such as OMP and TwIST algorithms. Compared with these two algorithms, the result of this paper shows that, the parallel Bregman algorithm needs shorter time, and thus is more convenient for real-time object reconstruction, which is important to people’s fast growing demand to information technology.
KEYWORDS: Light emitting diodes, Cameras, High speed cameras, Field programmable gate arrays, Signal processing, LED lighting, Imaging systems, 3D image processing, Image processing, CMOS cameras
Currently most domestic factories still manually detect machine arbors to decide if they meet industry standards. This method is costly, low efficient, and easy to misjudge the qualified arbors or miss the unqualified ones, thus seriously affects factories’ efficiency and credibility. In this paper, we design a specific high-speed camera system with auto adjustable ROI for machine arbor’s outline dimension measurement. The entire system includes an illumination part, a camera part, a mechanic structure part and a signal processing part based on FPGA. The system will help factories to realize automatic arbor measurement, and improve their efficiency and reduce their cost.
Compressed sensing (CS) is a new branch for information theory from the development of mathematical in 21st. CS
provides a state-of-art technique that we can reconstruct sparse signal from a very limited number of measurements.
In CS, reconstruct algorithm often need dense computation. The well-know algorithms like Basis Pursuit (BP) or
Matching Pursuit (MP) is not likely to implement in PCs in practice. In this paper, we consider to use GPU (Graphic
Processing Unit) and its large-scale computation ability to solve this problem. Based on the recently released NVIDIA
CUDA 6.0 Tool Kit and CUBLAS library we study the GPU implementation of Orthogonal Matching Pursuit (OMP), and
Two-Step Iterative Shrinkage algorithm (TwIST) implementing on GPU. The result shows that compared with CPU,
implementing those algorithms on GPU can get an obvious speed up without losing any accuracy.
KEYWORDS: Signal to noise ratio, Imaging systems, Principal component analysis, Compressive imaging, Low light level imaging, Image processing, Signal processing, Digital micromirror devices, Linear filtering, Image compression
In this paper, a compressive imaging architecture is used for ultra low-light-level imaging. In such a system,
features, instead of object pixels, are imaged onto a photocathode, and then magnified by an image intensifier.
By doing so, system measurement SNR is increased significantly. Therefore, the new system can image objects
at ultra low-ligh-level, while a conventional system has difficulty. PCA projection is used to collect feature
measurements in this work. Linear Wiener operator and nonlinear method based on FoE model are used to
reconstruct objects. Root mean square error (RMSE) is used to quantify system reconstruction quality.
A compressive imaging based APT (CI-APT) system is studied for FSO communication. Linear combinations
of object pixels, referred to as features, are measured. Then reconstructed objects are used for target locating.
Because it is implementation friendly, Hadamard projection is employed for CI-APT. Spatial domain andWavelet
domain OMP methods are studied for signal reconstruction. To demonstrate the idea, we use 64 randomly
selected Hadamard features to locate a 3 × 3 target in a 256 × 256 object. The averaged location error is less
than 2 pixels.
A real-time people counting system using ranging technology with human head-shoulder profile is discussed in
this work. To obtain the profile, the system is installed above an entrance/exit gate with vertically downward
view. Line structured light is used to detect the height of a person’s head-shoulder. Compared with imaging-
processing approaches, this method is cost-effective, computationally simple, and more accurate. The system
can also detect the walking direction of each person using a second structured illumination light source.
Precise online measurement of cable diameter and length will ensure stability of quality production and production speed. This paper describes a measuring device to measure cable length and diameter simultaneously, accomplish online production process control. The device consists mainly of a synchronous sampling part, calculation and control part. The synchronous sampling part consists of two parallel rollers one meter far from each other. A measuring band cinctures the rollers and move together with them. There are two electromagnetic chucks on the measuring band which are controlled by two photoelectric position switches to hold or release a cable, in order to make the cable move with measuring system synchronously. An optical encoder is connected to one of the rollers coaxially to measure cable length. For cable diameter measurement, two orthogonal CCD sensors are used. Accuracy of online diameter measurement is mainly affected by vibration of cable movement. In order to reduce the cable diameter measurement error caused by vibration, measuring system uses a mechanical damping device and high-speed CCD sensors which exposure time is up to microseconds. The calculation and control part of measuring device can filter, amplify and binarizate electrical signals from synchronous sampling part, then they are processed by microcontroller 8051 to complete cable length and diameter measurement. As well, the measuring device can set error limits and detect online whether cable length and diameter size are in default range , if not it would give corresponding alarm.
Compressed sensing or compressive sampling (CS) is a new framework for simultaneous data sampling and compression
which was proposed by Candes, Donoho, and Tao several years ago. Ever since the advent of a single-pixel camera, one
of the CS applications - compressive imaging (CI, also referred as feature-specific imaging) has aroused more interest of
numerous researchers. However, it is still a challenging problem to choose a simple and efficient measurement matrix in
such a hardware system, especially for large scale image. In this paper, we propose a new measurement matrix whose
rows are the odd rows of N order Hadamard matrix and discuss the validity of the matrix theoretically. The advantage of
the matrix is its universality and easy implementation in the optical domain owing to its integer-valued elements. In
addition, we demonstrate the validity of the matrix through the reconstruction of natural images using Orthogonal
Matching Pursuit (OMP) algorithm. Due to the limitation of the memory of the hardware system and personal computer
which is used to simulate the process, it is impossible to create such a large matrix that is used to conduct large scale
images. In order to solve the problem, the block-wise notion is introduced to conduct large scale images and the
experiments results present the validity of this method.
We describe a high-speed, high-resolution, and real-time scanning measurement system consisting of a linear laser, a
smart camera, a PC and the corresponding software. The smart camera with high-speed processing capability could
process the image of the object to be measured which is illuminated by the laser to get the data about the shape of the
object's cross-section profile in real time. We just need to transport the measured data rather than the huge number of
original image to PC for archiving or other application. By the relative motion between the system and the object, we can
get a series of data about the whole object's profile which can be reconstructed in the PC by corresponding application
software. The system was designed to be installed on the vehicles. With the moving of the vehicle we can get the shape
of the road.
This paper presents a new color image analysis approach which fuses several processing maps by Graph-Cuts algorithm
for Markov Random Field (MRF) in different color spaces. Recently, graph-based image analysis methods, such as
Graph-Cuts, have been achieved exciting results for approximate inference in Markov Random Field. One color image
can be represented in many color spaces, such as RGB, LAB and HSV, but existing Graph-Cuts approaches often
compute global energy function only in one color spaces and ignore that each color space has an interesting property for
certain applications respectively. This paper processes images in MRF and represents them in one MRF model firstly.
Then Graph-Cuts algorithms are used to process images in each color space and generate one map. Several processing
maps can be acquired from some color spaces. These maps are fused to get more reliable and accurate results. We select
stereo matching which can get depth maps from multi-view images to evaluate our image analysis approach. The
experiments herein reported in this paper illustrate the potential of this approach compared to existing Graph-Cuts
methods from processing results.
Electronic Rolling Shutter (ERS) was usually considered to be the shortcoming of low-cost and low-power CMOS image sensors, because ERS will cause distorted and/or blurred images when the target is moving. In this paper, we propose a new method to measure roller's velocity by using ERS's distortion as a response of moving target rather than a drawback. In the presented method, dimension expanding method is improved to generate a repeated isosceles triangle pattern as the reference object. 2D velocity is simultaneously calculated from the distorted isosceles triangle pattern. Some experimental results are given to clarify the feasibility of the proposed method.
A Number Plate Recognition System with a special linear CCD imager is presented in this paper. The system is fixed on the top of the roadway to obtain the images of the moving vehicles. With the linear image sequence, we have detected the position of Number Plate with corresponding algorithm and output 2-dimension sub-image of Number Plate by standard video for Optical Character Recognition (OCR). The 2-dimension image combined with the linear image sequence has higher transverse resolution (2048pixels) than that of the normal standard video (768pixels). Moreover, even images and accurate location of the characters can be gained with the linear CCD imager. The system is based on FPGA technology, and it can detect the vehicles Real-timely and recognize Number plate.
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.