A MACH Filter-Based Reconstruction-Free Target Detector and Tracker for Compressive Sensing Cameras

Compressive sensing cameras hold the promise of cost-effective hardware, lower data rates, and improved video quality, particularly outside the visible spectrum. However, these improvements involve significant computational cost, as sensor output must be reconstructed in order to form an image viewable by a human. This paper describes a prototype automated detection and tracking system using a compressive sensing camera that does not rely on computationally costly image reconstructions. It operates on raw sensor data for an approximately ten-fold improvement in computation time over a comparable reconstruct-then-track algorithm. The detector is successful at a sensing rate of 0.3, comparable to that required for high-quality image reconstructions. If initialized with the location of a target, the tracker holds the target at a sensing rate of 0.005, below the boundary where reconstruction breaks down. These results show not only that direct tracking from compressive cameras is possible, but also give support to the pursuit of direct inference from compressive sensors of all types.

Scale Distorted Target Recognition Based on Machine Vision

<div> <p class="9"><span>Machine vision means to carry out measurement or judgment with machine instead of human eyes. In the field of target recognition, optical correlation technology is a main way to realize machine vision. Targets can be recognized and located with high precision</span><span> </span><span>taking advantage of optoelectronic hybrid joint </span><span>transform</span><span> correlator (OHJTC). However, when scale or angular distortion of the detected target exists relative to the reference template, the intensity of correlation peaks will decrease to a great extent, which restricted the recognition results greatly. In this paper, the development and principle of maximum average correlation height (MACH)</span><span> </span><span>algorithm is introduced. Through amounts of experiments, t</span><span>he control</span><span>ling</span><span> parameters</span><span> </span><span>of the synthesized filter</span><span> are optimized</span><span>, which makes MACH filter suppress background noise and widen recognition range of targets</span><span>. To show the feasibility of this algorithm, simulative and optical experiments of the improved MACH filter are carried out. As an example, the recognition results of a fighter target in sky</span><span> </span><span>are given, which shows </span><span>the </span><span>scale</span><span> </span><span>distortion tolerance can reach up </span><span>to </span><span>±</span><span>23%</span><span>. The actual effect of the improved MACH filter algorithm has been confirmed very well</span><span>.</span><span> <o:p></o:p></span></p> </div>

Spatial Distorted Target Recognition Based on Improved Maximum Average Correlation Height Algorithm

Spatial distorted target is very hard to be recognized for complexity and variety of targets, which has restricted the development of pattern recognition technology to a great extent. Joint transform correlator is one of the key equipments to detect and recognize distorted targets. The appearance and development of maximum average correlation height (MACH) algorithm is introduced in this paper. Based on the principle the algorithm and experimental analysis, an improved maximum average correlation height algorithm fit for joint transform correlator is proposed, which has powerful capability of suppressing background noise and widening distortion tolerance. Target images with different shapes including scale or angular distortion constitute MACH filter in frequency domain, which is projected to space domain as reference template including varieties of attitude. To show the feasibility of the algorithm, an airplane with angular distortion in sky is processed by MACH filter as an example. Simulation and optical experimental results are given in this paper. The experiments show the angular distortion tolerance can reach up to 15 degrees. The actual effect of the improved MACH filter algorithm is confirmed very well.

Distorted Target Recognition Based on Prewitt Operator Combined with MACH Filter

The distorted (scale and rotation etc.) target images and the training images are edge extracted by Prewitt operator. The Maximum Average Correlation Height filter is synthesized with the training images. The low frequency information of the distorted target images and the filter is enhanced. Then the edge extracted distorted target image is filtered by the Maximum Average correlation Height filter. Thereby, the distortion tolerance of the Maximum Average Correlation Height filter is expanded. By this method, when the space edge extraction combines with frequency domain filtering, the scale distortion tolerance is 0.76~1.56 times; the rotation distortion tolerance can reach up to 80 degrees. For the targets exceeding the distortion tolerance range, the intensity of correlation peak becomes weaker and weaker, but some of the targets can still be identified.