S. Radha

Non-iterative Pseudo Inverse Based Recovery Algorithm (NIPIRA) for Compressively Sensed Images and Videos

Recent development in data compression emphasizes compressed sensing technique as a widely applied one for compression and reconstruction of images and videos by projecting the pixel values into smaller dimensional measurements. These compressed measurements are reconstructed at the receiver using suitable reconstruction algorithms, generally the greedy algorithms. Greedy algorithms are time consuming and complex processes, giving rise to a trade-off between reconstruction performance and algorithmic performance. This work proposes a non-iterative method, non-iterative pseudo inverse based recovery algorithm (NIPIRA), for reconstruction of compressively sensed images and videos that exhibits small complexity and time requirement along with preservation of reconstruction quality. Mathematical proofs for NIPIRA’s accuracy and optimality provide additional theoretical support to the algorithm. NIPIRA gives a minimum PSNR of 32 dB for very few measurements, accuracy of above 97 and 92% decrease in elapsed time compared with other iterative algorithms. The complexity of NIPIRA is \(O(MN)\) which is \(s\) times less than OMP and StOMP.

Efficient compressed sensing based object detection system for video surveillance application in WMSN

Limited memory, energy and bandwidth are the major constraints in wireless visual sensor network (WVSN). Video surveillance applications in WVSN attracts a lot of attention in recent years which requires high detection accuracy and increased network lifetime that can be achieved by reducing the energy consumption in the sensor nodes. Compressed sensing (CS) based background subtraction plays a significant role in video surveillance application for detecting the presence of anomaly with reduced complexity and energy. This paper presents a system based on CS for single and multi object detection that can detect the presence of an anomaly with higher detection accuracy and minimum energy. A novel and efficient mean measurement differencing approach with adaptive threshold strategy is proposed for detection of the objects with less number of measurements, thereby reducing transmission energy. The performance of the system is evaluated in terms of detection accuracy, transmission energy and network lifetime. Furthermore, the proposed approach is compared with the conventional background subtraction approach. The simulation results show that the proposed approach yields better detection accuracy with 90% reduction in samples compared to conventional approach.

Efficient compressed sensing-based security approach for video surveillance application in wireless multimedia sensor networks

Video surveillance application in wireless multimedia sensor networks (WMSNs) require that the captured video must be transmitted in a secured manner to the monitoring site. A compressed sensing (CS)-based security mechanism is proposed in which the security keys are generated from the measurement matrix elements for protecting the users identity. The security keys are applied for protecting the video from being reconstructed by the attacker. The proposed framework is tested in real time using a WMSN testbed and the parameters such as memory footprint, security processing overhead, communication overhead, energy consumption, and packet loss are evaluated to demonstrate the effectiveness of the proposed security framework. The results showed that the proposed security mechanism has 92% less storage complexity compared to an existing CS-based security mechanism. The energy consumed for transmitting the secured measurements is 53% less when compared to raw frame transmission.

Web Enabled Plant Disease Detection System for Agricultural Applications Using WMSN

Plant disease detection attracts significant attention in the field of agriculture where image based disease detection plays an important role. To improve the yield of plants, it is necessary to detect the onset of diseases in plants and advice the farmers to act based on the suggestions. In this paper, a novel web enabled disease detection system (WEDDS) based on compressed sensing (CS) is proposed to detect and classify the diseases in leaves. Statistical based thresholding strategy is proposed for segmentation of the diseased leaf. CS measurements of the segmented leaf are uploaded to the cloud to reduce the storage complexity. At the monitoring site, the measurements are retrieved and the features are extracted from the reconstructed segmented image. The analysis and classification is done using support vector machine classifier. The performance of the proposed WEDDS has been evaluated in terms of accuracy and is compared with the existing techniques. The WEDDS was also evaluated experimentally using Raspberry pi 3 board. The results show that the proposed method provides an overall detection accuracy of 98.5% and classification accuracy of 98.4%.

Energy efficient surveillance system using WVSN with reweighted sampling in modified fast Haar wavelet transform domain

Wireless visual sensor network (WVSN) consists of a large number of nodes that are capable of acquiring, compressing and transmitting images. Surveillance becomes a vital application area of WVSN as they can be deployed in various environments to monitor and collect information. The lifetime of the nodes in the network depends on the energy consumption. Hence in this paper, block compressed sensing (BCS) based image transmission technique that utilizes Energy based Reweighted sampling (ERWS) in Modified Fast Haar Wavelet Transform (MFHWT) domain is proposed to reduce energy consumption considerably. Sparse binary random matrix is used to acquire CS measurements in MFHWT domain. In addition, the proposed technique also maintains the image quality. The developed algorithm is applied and tested for a car parking lot monitoring system. It is evident from the simulation results that the proposed method achieves better PSNR values even for fewer measurements. Experimental analysis is performed using Atmega 128 processor of Mica mote in WinAVR by Atmel. The proposed method has approximately 85.5% lesser energy consumption than other Compressed Sensing (CS) methods. Lossless Entropy Coding is applied to the ERWS measurements and considerable reduction in number of transmitted bits is also achieved. The algorithm has also been tested in WINGZ mote in real time.

Noise performance of non-iterative compressed sensing based recovery algorithm: surveillance applications

Compressed sensing has been of great interest in signal compression since it promises higher compression level and ease in usage. It is widely used in signal processing domain for compression and reconstruction of various signals including electrical signals, images, videos, etc. The concept of compressed sensing can be applied suitably for surveillance videos since voluminous video quantities can be significantly compressed and retrieved perfectly. The surveillance videos are prone to various noises like impulse noise, quantization noise, multiplicative noise, etc., that raise as hindrance to high quality video reconstruction. Thus, non-iterative compressed sensing based recovery algorithm is proposed that recovers the surveillance videos with higher perfection in the presence of various noises. The algorithm uses augmented matrix as sensing matrix and hence avoids iterations leading to commendable reduction in runtime. The signal to noise ratio obtained using the proposed algorithm is ~39 dB which is greater than any other existing noise removing CS recovery algorithms like OMP-CV, TMSBL, etc. High speed recovery is made possible due to the absence of iterations. Accuracy and structural similarity obtained are nearly 98% and 95% respectively. The algorithm is robust to various noise levels and the hardware implementation shows that the algorithm is simple enough to be used in hardware of lower specifications. These results ensure NIPIRA as a best suitor for real time surveillance video reconstruction even in the presence of noise.

Design of PDMS membrane for CTC separation

Circulating tumorcells(CTC) are the primary tumor cells which contains the key information about the cancer. These CTCs can be separated from the peripheral blood in microfluidics, where a PDMS(polydimethylsiloxane) membrane can be designed for CTC detection. The PDMS membrane is designed with cylindrical pores for the detection of Circulating TumorCells(CTC) from the peripheral blood samples. The membrane consisting of cylindrical pores of size 9μm and the distance between the pores are designed to be (15μm) are simulated and analysed for the separation of CTCs.

Efficient Anomaly Detection System for Video Surveillance Application in WVSN with Particle Swarm Optimization

Wireless sensor networks consist of several tiny low cost sensor nodes that are deployed for many applications such as military, civil, industrial, healthcare, home automation, etc. Recent technological developments have enabled the use of wireless visual sensor networks (WVSNs) for sensitive applications such as video surveillance and monitoring applications. Limited memory, energy and bandwidth are the major constraints in WVSN that can be simplified by the use of compressed sensing (CS), which asserts that sparse signals can be reconstructed from very few measurements. CS a computational intelligence solution is about acquiring and recovering the signal in the most efficient manner possible using incoherent projection basis. In the case of video surveillance applications, the entire video may not be useful hence, with the help of efficient algorithms the presence of the anomalies can be detected and transmitted to help user at the monitoring site to take necessary action. In this chapter, particle swarm optimization (PSO) based efficient anomaly detection system (EADS) is proposed which will detect the presence of anomalies and transmit the required measurements via TelosB nodes to the network operator. This system adopts the concept of CS to obtain the compressive measurements so that the object detection algorithm can be applied to the measurements rather than samples. PSO is employed for optimizing compressive measurements while a mean based measurement differencing approach is used for detecting the object. This proposed efficient system has the intelligence of detecting targets with fewer measurements and transmit the required compressive measurements for reconstruction with less energy, thereby increasing the network lifetime. PSO is used to optimize the transmission distance with minimum number of hops towards destination, to achieve reduced energy consumption. However, the lifetime of the network is still bounded by batteries, the sole source of energy in WVSNs. Alternative energy utilization can be effectively included to recharge the batteries on-board and extend the lifetime of the network. Solar energy harvesting forms an effective resource due to its ambient presence. Hence, solar energy harvester is incorporated in the proposed EADS to extend its lifetime.

Separation of bio-particles in micro fluidic device

Conventionally separation of bio-particle from blood is a long term process. Different methods are involved in separation of bio-particles. Bio-particles like bacteria, WBC, RBC are separated in a micro fluidic device by using pillars and applying different types of fields. In this paper, different shapes of pillars are analysed for efficient separation of bio-particles without applying different types of field, which pays way for the development of bio-particle separation filter. The overall length of the channel is 1cm and the size of the pillars is 12×15×8 μm.

Design of a bio-filter for particle separation

Cell separation and the diagnosis is an important step in the medical or Biological areas of research. The cell separation can be done by various techniques. A design is built with an array of micro-pillars where the micro-pillars acts as a filters which is used for the separation of spherical and non-spherical particles without the use of any external fields. The micro-pillars are designed in I-shape rather than a conventional shapeto separate plasma from blood where the diameter and height of each pillar is 12μm and 15μm respectively and the distance between the pillars are 8μm at the middle and 2μm at the edges. This paper discuss about the design of filter for particles separation without applying any external fields, consuming lower bio samples and lesser processing timein biomedical application.