1. Introduction1.1. Qualifications2. Approach3. Questions3.1. Stabilization3.1.1 Will an affine model work for image background stabilization?3.1.2 Are image enhancements necessary for stabilization (histogram equalization, edge enhancements, etc.)?3.2. Periodic motion3.2.1 Will temporal filtering expose the periodic nature of the radar rotation? Is the period accurately measured?3.2.2 Will sequence alignment detect the rotation of the radar? Is the period accurately measured?3.2.3 Does one method outperform the other?3.2.4 Can prior knowledge of the rotation periods of radars improve the performance?3.3. Analysis3.3.1 Can the pixel width of the radar be estimated?3.3.2 If there are known radars of specific physical lengths rotate at specific periods, can the distance of the target be estimated?3.3.3 Can it be determined the radar is facing the camera or observer system within 180 degree phase? 3.4. Noise3.4.1 Are there false positives when people walk near the radar? If so, are there methods for minimizing false positives?3.4.2 Can multiple radars be detected in one video sequence? 4. Experiment Resources4.1. Data Collection4.2. Software Tools5. MilestonesReferencesRadar Antenna Detection and Analysis for Autonomous SEI Collection Kris Gibson Department of Electrical Computer Engineering University of California, San Diego [email protected] Abstract The ability of an unmanned system to visually determine the direction a radar antenna is facing can significantly improve the SEI collection process and provide opportunities for autonomous collection. I will investigate computer vision based methods for detecting and analyzing periodic motion of spinning radar in the maritime domain. 1. Introduction SEI collection is an important asset for information dominance. “SEI provides a reliable, long-range, all-weather positive target identification capability against seaborne platforms and land-based systems that emit radar signals.” [1]. There are a multitude of research journals that address the electromagnetic signal parameter realm of SEI. However there is little research in automating the SEI collection process. The SEI collection procedure currently requires a human-in-the-loop. The operator (human-in-the-loop) is required to point the SEI collection antenna at a ship’s spinning radar using optics bore-sighted with the collection antenna. The operator (sometimes a second operator) configures the SEI collection tuning parameters while the optics are manually positioned to follow the ship radar. An SEI collection occurs when the operator confirms the position of the radar antenna is directing its beam at the collection antenna when the SEI system signifies a signal has been received. The collection is complete after the operator reads the name of the ship, confirms the radar signal is indeed from the ship, and acquires distinguishing other features (color, hull type, etc.) The procedure the operator employs is an important step for providing ground truth. One of the four key core concepts for an SEI system design is “Providing ground truth (the correct identification of the emitters being evaluated) for the naming of the clusters and evaluation of the clustering process.” [2]. This is a complex, error-prone, and time-consuming procedure which could be significantly optimized through automation. The automation is addressed by researching methods for radar angular position and periodicity verification. For most collection sites, this step alone will cut down the need of two operators down to one. The science addressing this vision system concept can also be extended to provide cueing for ship detection, tracking, and identification. Another reason for autonomy is collection in hostile environments. SEI collection using the current procedures is not a safe option because of the risk of placing human operators in hostile environments. This risk limits the ability to monitor vessels of interest (VOI). A VOI can avoid intelligence collections by staying within the hostile environments. An autonomous SEI collection system would best support this scenario and could be tremendously advantageous to the DoD mission of MDA. 1.1. Qualifications Kris Gibson is currently in the Master of Engineering program at UCSD. He has taken Computer Vision I (CSE 252A), Image Processing (CSE 166), and is currently taking Digital Image Processing (ECE 253A). Kris has experience as a lead software engineer in the Electro-Optics Surveillance (EOS) group at SPAWAR Systems Center Pacific. He was the EOS subject matter expert (SME) in video and still imagery. Kris has supported the development and deployment of surveillance systems which have been deployed around the world. As the lead software engineer and image enhancement Project Engineer, Kris has designed, built, and deployed many key software modules including: image and video enhancement capabilities for real time and post processing applications to provide solutions to counter marine environment atmospheric affects; and image and video capture for intelligence analysts.2. Approach There are three phases to address all four previously mentioned problem areas: 1) Radar detection and position determination 2) Ship name reading 3) Locating radar and ship name. The first phase is addressed in this proposal. Periodic motion of objects is a widely addressed problem in the computer vision community. Detecting locally periodic motion of humans, animals, and vehicles has been demonstrated [3]. This approach assumes the video sequences are stabile and the object is already tracked. In contrast, the detection of periodic motion in humans in complex scenes has been demonstrated without the precondition of stabilization and segmentation using approximate sequence alignment [4]. My approach will also require the relaxed assumption that the image sequence has not been preconditioned such as stabilized, tracked, and segmented. The first step in developing a model is to research methods for stabilization and enhancement. The idea of sequence alignment will then play a key factor in establishing a method for detecting periodicity. 3. Questions The following numbered items outline questions to address while investigating the first phase which is radar detection and position determination. 3.1. Stabilization 3.1.1 Will an affine model work for image background stabilization? 3.1.2 Are image enhancements necessary for