Distributed imaging using an array of compressive cameras

Ke, Jun; Shankar, Premchandra M.; Neifeld, Mark A.

doi:10.1016/j.optcom.2008.09.083

Cited by 9 publications

(7 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Increasing DRAM memory in the D4100 controller board, such that all the patterns could be loaded prior to image capture, would significantly speed up data acquisition to a 1% sampling runtime of 1.4 min or less. Even faster acquisition could be achieved, and dynamic range improved, by multiplexing parts of the spatially filtered signals onto multiple parallel detectors, as proposed by Ke et al [22], providing a continuum between conventional focal planes and compressive imagers. In the longer term, ASIC image sensors with high pixel counts and programmable on-chip signal aggregation will be able to integrate the pattern encoding directly into the image sensor itself, eliminating the need for an external SLM and enabling CI systems to function with high-performance image formation optics [17,18].…”

Section: Discussionmentioning

confidence: 99%

Characterization of a compressive imaging system using laboratory and natural light scenes

Olivas

Rachlin

et al. 2013

Appl. Opt.

View full text Add to dashboard Cite

Compressive imagers acquire images, or other optical scene information, by a series of spatially filtered intensity measurements, where the total number of measurements required depends on the desired image quality. Compressive imaging (CI) offers a versatile approach to optical sensing which can improve size, weight, and performance (SWaP) for multispectral imaging or feature-based optical sensing. Here we report the first (to our knowledge) systematic performance comparison of a CI system to a conventional focal plane imager for binary, grayscale, and natural light (visible color and infrared) scenes. We generate 1024 × 1024 images from a range of measurements (0.1%-100%) acquired using digital (Hadamard), grayscale (discrete cosine transform), and random (Noiselet) CI basis sets. Comparing the outcome of the compressive images to conventionally acquired images, each made using 1% of full sampling, we conclude that the Hadamard Transform offered the best performance and yielded images with comparable aesthetic quality and slightly higher spatial resolution than conventionally acquired images.

show abstract

Section: Discussionmentioning

confidence: 99%

Characterization of a compressive imaging system using laboratory and natural light scenes

Olivas

Rachlin

et al. 2013

Appl. Opt.

View full text Add to dashboard Cite

show abstract

“…A consequence, however, is that in each DW-OSH measurement the object signal is reduced by half because of the reduced exposure. The detector needs to work with a larger bandwidth, leading to more detector noise [10,11]. Specifically, if we assume the detector exposure time in SW-OSH for one measurement is T 0 , then the noise energy is proportional to σ 2 0 ∕T 0 , where σ 2 0 is the noise energy per bandwidth.…”

Section: A Osh System With a Source Working At Two Wavelengthsmentioning

confidence: 99%

Depth resolution enhancement in optical scanning holography with a dual-wavelength laser source

Poon

Lam

2011

Appl. Opt.

Self Cite

View full text Add to dashboard Cite

In this paper, we use two point sources to analyze the depth resolution of an optical scanning holography (OSH) system with a single-wavelength source. A dual-wavelength source is then employed to improve it, where this dual-wavelength OSH (DW-OSH) system is modeled with a linear system of equations. Object sectioning in DW-OSH is obtained with the Fourier domain conjugate gradient method. Simulation results show that, with the two source wavelengths at 543 nm and 633 nm, a depth resolution at 2.5 μm can be achieved. Furthermore, an OSH system emulator is provided to demonstrate the performance of DW-OSH compared with a conventional OSH system.

show abstract

“…Distributed optimization has been receiving increasing attention in the last years [1], [2], [3], [4], [5] due to its applications in diverse multi-agent frameworks, ranging from detection and estimation over sensor networks [6], [7] to compressed sensing [8] and medical imaging [9]. Modern networked technologies have demonstrated that distributed systems of interconnected and low-power units can efficiently replace a single, powerful centralized processor for tasks like, e.g,, monitoring, tracking, localization, and imaging [9], [10], [11], [12], [13]. In some cases, networked systems are used only to acquire data, and optimization is performed by a single data fusion center.…”

Section: Introductionmentioning

confidence: 99%

Randomized Algorithms for Distributed Nonlinear Optimization Under Sparsity Constraints

Ravazzi

Fosson

Magli

2016

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

Distributed optimization in multi-agent systems under sparsity constraints has recently received a lot of attention. In this paper, we consider the in-network minimization of a continuously differentiable nonlinear function which is a combination of local agent objective functions subject to sparsity constraints on the variables. A crucial issue of in-network optimization is the handling of the communications, which may be expensive. This calls for efficient algorithms, that are able to reduce the number of required communication links and transmitted messages. To this end, we focus on asynchronous and randomized distributed techniques. Based on consensus techniques and iterative hard thresholding methods, we propose three methods that attempt to minimize the given function, promoting sparsity of the solution: asynchronous hard thresholding (AHT), broadcast hard thresholding (BHT), and gossip hard thresholding (GHT). Although similar in many aspects, it is difficult to obtain a unified analysis for the proposed algorithms. Specifically, we theoretically prove the convergence and characterize the limit points of AHT in regular networks under some proper assumptions on the functions to be minimized. For BHT and GHT, instead, we characterize the fixed points of the maps that rule their dynamics in terms of stationary points of original problem. Finally, we illustrate the implementation of our techniques in compressed sensing and present several numerical results on performance and number of transmissions required for convergence.

show abstract

Distributed imaging using an array of compressive cameras

Cited by 9 publications

References 12 publications

Characterization of a compressive imaging system using laboratory and natural light scenes

Characterization of a compressive imaging system using laboratory and natural light scenes

Depth resolution enhancement in optical scanning holography with a dual-wavelength laser source

Randomized Algorithms for Distributed Nonlinear Optimization Under Sparsity Constraints

Contact Info

Product

Resources

About