Motion blur of an image is a common phenomenon that occurs while taking a photograph due to the relative movement of the object and an image acquiring device. It is essential to detect this phenomenon of blurring of images in many applications such as information retrieval. This paper proposes a novel local blur detection technique, and it performs better than the existing works. This technique mainly uses Radon transform and Laplacian of Gaussian on the local neighborhood around each pixel to estimate blur information. Additionally, two new weight functions are introduced based on local geodesic distance and local variance. It is shown that these functions play a significant role in segregating blur and non-blurred parts. Simulation results validate the correctness and accuracy by testing the proposed algorithm on some challenging images with similar color information in the foreground and background. Various quantitative performance measures have determined the superiority of the proposed method.

Airborne surveillance systems have multiple sensors and communication links on board a suitable platform. They work in a cohesive manner to provide effective surveillance over the region of interest. The performance proving of such a system is challenging and requires flight trails extending over years. The test results often have to be interpreted using statistical analysis of the flight test data. An efficient way is to carefully design the flight test profiles such that enough samples can be collected during the test and multiple requirements can be tested in a single sortie. Such meticulous test strategies where both own ship platform and test targets are moving with high dynamics call for software based tool for planning of test sorties and the test points. Flight Planning Tool (FPT) plays an important role in pre-flight stage during developmental trials for analysis of the MOEs and MOPs of overall system and of various on-board sensors of an airborne multi-sensor system. The FPT provides statistical & graphical analysis for sensor behaviour for various scenarios (flight trials) before actual flight test is conducted. It provides prior information on number of valid samples for sensor testing during flight trials. In addition, the tool aids in assessing number of profiles to be flown for proving each MOE. The profiles can also be optimised such that valid samples are collected for evaluation.

Airborne Electro-Optic /Infra-Red (EO/IR) system is the major workhorse for Maritime Surveillance and Search & Rescue (SAR) for object detection, recognition and identification from an airborne platform. Often, the EO/IR system is integrated as part of mission suite, which includes other sensors like Maritime Patrol Radar (MPR), Automatic Identification System (AIS), Communication Support Measures (CSM) etc. along with multiple Line of Sight (LOS) and Satellite links. This paper brings out the development of such a compact airborne HD EO/IR system along with its tight integration to a mission suite that is primarily intended at maritime surveillance and SAR roles from a Maritime Patrol Aircraft (MPA). This paper also describes the software integration of EO/IR with Mission Management System (MMS) and Multifunctional Tactical Console (MTC) achieving efficient and effective exploitation of the EO/IR system. The details of various user interfaces of MTC, command and control operations related to EO/IR along with map based operations are also brought out. Finally, we demonstrate such an integration in a lab based set up.

In this paper, we propose a novel multi modal object segmentation model based on user inputs. In the proposed strategy, two new fidelity terms are introduced that comprise of region marker points and geodesic distance. Here, region markers points help to select dominant intensity of the object. And, geodesic distance help in confining converging contour to remain around the boundary in the case of similar intensity values in the object and background. To minimise the proposed functional, we derived Euler Lagrange equation and solved it using finite difference scheme. Moreover, to evaluate the performance with other existing methods we implemented it in MATLAB and compared the outcomes by determining various performance measures. The experimental results shows that the proposed method outperforms the traditional selective segmentation active contour model in terms of accuracy and robustness.

Mission and Data Management System (MDMS) is the brain of any airborne surveillance system which includes mission planning and loading, provision of loading mission parameter data, command & control of different sensors and sub sensors, state management, periodic health monitoring, supply of discrete timing data, streaming of navigation data, integration of surveillance data from sensors like Radar, Identification Friend or Foe (IFF) and Automatic Identification System (AIS), providing decision aids, mission data, video, audio recording and retrieval. This paper provides a framework for integrating multitude of sensors for maritime surveillance using Mission Management System (MMS) and Multifunctional Tactical Console (MTC), which are constituent components of MDMS, in an integration test bed. The tight integration of various sensors to MMS will enhance the capability of the sensors as an integrated suite and the output from the various sensors are fused for enhancing the situational awareness. Such an architecture will enable mission ready integrated command and control of various sensors with high availability.

We propose a Constant False Alarm Rate (CFAR) algorithm called Extended Target Cell-Averaging-CFAR (ET-CA-CFAR) for extended target detection. We investigate its performance in Medium Resolution Radars (MRR), under Gaussian clutter. Extended target detection has been widely studied in High Resolution Radars (HRR). However, HRRs require a first stage detector to confirm the presence of a target. The first stage detections are usually performed using a Low Resolution Radars (LRR). In this paper, we bring out the advantages of using MRR with ET-CA-CFAR against LRRs. We derive closed form expressions for probability of detection and false alarm for ET-CA-CFAR under Gaussian clutter and analytically show the detection gains over traditional CA-CFAR. Our approach has lower computational complexity compared to HRRs and improved detection performance compared to LRRs. Thus, our approach can be used to improve range performance of small aperture surveillance radars, specifically airborne surveillance radars with size limitations.

Airborne maritime surveillance plays an important role in safeguarding the coastline and exclusive economic zones. A multi-functional Maritime Patrol Radar (MPR), which can provide 360 degree azimuth coverage and operate in various modes like sea surface surveillance, Synthetic Aperture Radar (SAR), Inverse SAR (ISAR), Search-and-Rescue, Weather etc., is crucial for providing an effective maritime situational awareness. The performance of such systems are usually limited by sea clutter and aperture constraints imposed by the aircraft. Unlike the older generation mechanical radar, MPR based on Active Electronically Scanned Array (AESA) can overcome these limitations with its capability in instantaneous beam steering, spacetime agility etc. This paper performs an extensive simulation study on sea-clutter, derives the optimal system parameters and defines a prototype Integrated X-band Active Phased Array Radar (IXAPAR). Each building block of IXAPAR has been designed using state-of-the-art technology. Finally, we report the performance results of the IXAPAR including opportune target detection, with various advanced digital domain processing like Extended Target CA-CFAR, digital monopulsing, simultaneous multi-direction looks using digital beamforming and jammer cancellation. The prototype IXAPAR can be easily scaled up based on various airborne and ground based applications.

In this article, a novel method for the interactive segmentation of the object and background which has multiple piecewise homogeneous intensities has been introduced. We have proposed a new energy function based on intensity points selected from multiple homogeneous regions in the object and background. To minimize the derived energy function, we use the calculus of variation method and transformed it to PDE. The derived PDE has been solved using an additive operator splitting (AOS) method. Simulation results validate the correctness and accuracy of the proposed method. The performance is assessed by testing the algorithm on synthetic images and results are compared with state of the art methods using Jaccard’s similarity index.

Selective image segmentation extracts object of interest from an image based on user input. There are many variational methods which are effective in segmenting object which has uniform intensity. But many times object with multiple intensities needs to extracted. In this paper, we propose an active contour based algorithm for selective segmentation of objects in vector valued image. Here, we introduced new energy function by adding new geometric constraints based fidelity terms from local region of the object for vector valued images. The model minimizes new functional over the length of the contour along various parts of the object in different component images. Experimental results shows that the proposed method is effective in segmenting object having multi intensities. Performance of the proposed method is determined using Jaccard’s similarity index for all methods.

Active electronically scanned antenna (AESA) based radars imbibe the desirable feature of 'graceful degradation'. Such radars use miniaturised transmit-receive (TR) modules and a failure of few modules does not lead to failure of the mission. For example, in AESA-based ground MTI radar, failure of a few modules does not affect the array performance. In such a case, the static ground clutter is centred on zero frequency does not have a motion dependent Doppler spread. However, in airborne AESA radars, the ground clutter has an angle dependent Doppler frequency due to the platform motion and clutter leaking in through antenna side-lobes. Hence, the antenna side lobe levels dictate the side lobe clutter against which target detection is to be performed. The detection performance is governed by the signal to interference plus noise ratio (SINR). For Airborne surveillance radar the effect of random and systematic failures of TR modules and their effect on SINR is characterised. It is shown that single channel processing does not effectively provide the graceful degradation feature as the SINR loss due to failures is significant. However, the effect of systematic failure on SINR loss is less as compared to random failures. An effective scheme for feeding the array is also proposed.

In this paper, an object blur detection and deblurring technique is proposed to restore multi-directional motion blurred objects in a single image. We have proposed local blur angle detection method based on Radon transform (RT) and Laplacian of Gaussian (LoG). While capturing the images, motion blur occurs mainly due to either movement of the objects or movement of the camera. Here, we have focused to restore the objects which has been blurred by motion of the objects. The estimation of likely blur direction is calculated in the blurred image using RT and gradient operators. To detect blur angle locally at each pixel, the new local blur angle estimator using RT and LoG has been developed. Numerical experiments have been carried out for the proposed method, and the results are compared with the state-of-the-art methods.

Space-time adaptive processing (STAP) based techniques has gained popularity due to their ability to adaptively cancel the interference in the angle Doppler domain. The major contributor of interference in airborne radar is ground clutter. For side-looking airborne radar, the ground clutter is localized to the diagonal in the angle Doppler domain. However, for non-side-looking airborne radar, the clutter exhibits higher rank and takes the shape of an ellipse. Also, practical schemes derive spatial channels from sub-arrays for dimensionality reduction. This paper characterizes the sub-array based STAP algorithm performance of a multistage Wiener filter for non-side-looking active electronically scanned array based airborne radar.

Inverse Synthetic Aperture Radar (ISAR) is used to image objects such as aircrafts and ships by the virtue of their rotation. The high resolution in range is obtained by large bandwidth pulses and similarly in the azimuth, by the rotational motion of the object. The rotational motion produces a Doppler shift, measured by the phase shift produced on a sequence of coherent pulses. Hence the azimuth resolution is determined by the Doppler discrimination using spectral analysis. In this paper, we propose an analysis technique for classification of ships, based on root-Multiple Signal Classification (MUSIC) algorithm, which betters the performance provided by the conventional Fast Fourier Transform (FFT) technique for ISAR imaging. This technique belongs to a class of super-resolution algorithms which also provides a resolution improved target classification.

Space-time adaptive processing (STAP) is essential for airborne radar due to their ability to detect targets under stressful interference scenario. The curse of dimensionality in STAP is overcome by adopting partially adaptive processing schemes. Joint Doman Localization (JDL) is a popular dimension reduced technique that adaptively nulls the interference in angleDoppler domain. This paper characterizes the performance of the JDL algorithm for real-world effects like sub-array beam-forming, aircraft crab, antenna pattern errors and channel mismatches. In what follows, we show that JDL is a robust algorithm. Further, a block JDL (B-JDL) algorithm is proposed for faster search over angle-Doppler Space.

Synthetic Aperture Radars (SAR) provide high resolution ground images by utilizing the aircraft motion to synthesize a large aperture. The presence of residual phase error in the SAR imagescauses defocusing in the images. Autofocus algorithm compensates these phase errors to obtain a focused image. In most scenarios a priori information about the type of phase error is known which can be effectively utilized in autofocus techniques. The paper proposes a new autofocus algorithm which is based on Wiener filter theory and implements a multistage Wiener filter to compensate the phase error in SAR images.

Secondary surveillance radar (SSR) aids in the identification of aircrafts and often works in conjunction with the primary radar for better target awareness. An important module of the SSR system is the reply signal processor unit. One of the major issues in reply processing is the occurrence of garbled replies, which occurs due to the overlapping of replies. Overlapping occurs when range difference between the two targets is less than de-garble resolution (3045 m). In practical scenarios SSR replies are often garbled (e.g.: formation flights). So efficient de-garbling algorithms are required for providing the target identification. In this paper a new algorithm towards de-garbling of SSR replies is discussed. The information derived from the delta to sigma ratio channel (Delta/Sigma) and the Sigma (sum) channel of the monopulse receiver are effectively combined for separating the garbled replies