Real-time Detection Algorithm for Small Space Targets Based on Max-median Filter

Abstract

A real-time detection algorithm for small space targets was presented based on FPGA and DSP, and the corresponding hardware was designed. With this system, the Video image, whose frame frequency reaches 40 Hz and the size reaches 512×512 pixels, could be processed in real-time application. Firstly, Max-Median filter was realized in FPGA to depress noise to solve the problem of mass computation and high speed request in image pre-processing. The segmentation of the moving target was done with an improved dual image difference and background suppression to ensure that small moving targets were detected in a low SNR and complex background image. And the algorithm could ensure diminishing false-alarm rate and maintaining detection probability. Practical application approved that the system perfectly meets the requirements of the real-time dim target detection and recognition in deep space.

Space Object Detection in Video Satellite Images Using Motion Information

“…The difference between (23) and (24) is the variable coefficient a xy based on detection of the previous frame. a xy is initially set to be a little greater than 1 and reset to the initial value when the gray level at x, y becomes large again (greater than 150).…”

“…Therefore, there is also a large amount of computation and poor real-time performance. A real-time detection algorithm based on FPGA and DSP for small space objects was presented by [23], but the instability and motion of observation platform were not considered. In addition, image sequences used to validate the algorithm in the literature above are from a ground photoelectricity telescope in [15,17,18,[21][22][23] and generated by simulation in [19,20]; the characteristic of image sequence from space-based observation has not been considered.…”

“…A real-time detection algorithm based on FPGA and DSP for small space objects was presented by [23], but the instability and motion of observation platform were not considered. In addition, image sequences used to validate the algorithm in the literature above are from a ground photoelectricity telescope in [15,17,18,[21][22][23] and generated by simulation in [19,20]; the characteristic of image sequence from space-based observation has not been considered. This paper studies space object detection in a video satellite image with star image background, and a new detection algorithm based on motion information is proposed, which includes not only the known satellite attitude motion information but also the unknown object motion information.…”

Context-aware cross-level attention fusion network for infrared small target detection

“…In past years, traditional methods mainly designed hand-crafted features based on various assumptions. They usually include filter-based methods, 2 – 5 human visual system (HVS)-based methods, 6 – 9 and low-rank sparse matrix factorization (LRSMF)-based methods 10 – 12 Filter-based methods use different filter templates to suppress the backgrounds and enhance the targets.…”

Autonomous Streaming Space Objects Detection Based on a Remote Optical System

“…Such systems provide the integration of all processing modules -detection, filtering, segmentation, astrometric calibration, conversion of pixel values into angular coordinates, classification, etc. [13] Pipeline processing of a video stream in applied problems of orbital monitoring is studied [14]. Detection and extraction of mathematical properties of space objects in real-time by remote video systems automatic algorithms solve the problem of manual processing of a large volume of astrometric data and makes it possible to calculate orbital parameters under observation conditions [15,16].…”