KEYWORDS: Data modeling, Image restoration, Data privacy, Pose estimation, Imaging systems, Sensors, Object detection, 3D modeling, Systems modeling, Simulations
The progress of imaging technologies has raised significant concerns regarding privacy and data transmission. Traditional imaging systems capture high-resolution images containing more personal information than necessary, posing privacy risks and requiring substantial computational resources. To address this issue, we propose a semi-image-free single-pixel imaging framework for human detection and pose estimation. The proposed model reconstructs only essential image regions, specifically bounding boxes containing humans, utilizing one-dimensional signals, thereby enhancing privacy and reducing data processing requirements. Simulations on the Microsoft COCO dataset are performed, showing our method is robust to the background noise. Our method operates efficiently with a low sampling ratio of approximately 3.5%, highlighting its suitability for resource-constrained environments.
KEYWORDS: Image restoration, Light sources and illumination, Sampling rates, Simulations, 3D tracking, Image processing, Modulation, Video processing, Video, Projection systems
Single-pixel fast-moving object tracking technique has potential applications in autonomous driving. However, employing a single pixel detector to the dynamic object imaging is a challenging task due to severe motion blur. Existing motion compensation methods are computationally expensive since they require the reconstruction of a series of images corresponding to the object moving process at different time intervals. In this paper, we propose a novel method that only needs to reconstruct one image of the object during the whole moving process, thus significantly reducing the computational cost. We achieve ultra-efficient simultaneously tracking and imaging of a moving object by designing novel illumination patterns. With our designed patterns, there is a mathematical similarity between the single-pixel imaging model and the low-order moment of the image. Thus, the detected light intensity can be directly used as the low-order moment value to calculate the position of the object. This work only needs two patterns for object localization in each frame, which is the current highest positioning frame rate in the single-pixel imaging field. For a 128×128 pixels object, with the employment of the Compressed Sensing (CS) and Total Variation Augmented Lagrange Alternative Direction Algorithm (TVAL3), the proposed new method can reconstruct a high-quality image with a sampling rate of 12.2% without any priori motion information. The proposed scheme paves the way for the development of fast-moving object imaging.
Single-Pixel Imaging (SPI) techniques enable the reconstruction of an image scene utilizing multiple spatially modulated light patterns and the corresponding measurements from a single-pixel detector. Since in SPI, the acquisition time scales quadratically with the image resolution. A high-resolution image reconstruction suffers from a slow reconstruction speed. This work proposes a super-resolution single-pixel imaging methodology based on generative adversarial networks (GAN-SRSPI). A low-resolution (N×N) image is reconstructed and then super-resolved to obtain a high-resolution (4N× 4N) image. The 4×super-resolution leads to an overall sampling rate of 6.25% (1/16). The previous work only minimizes the mean squared error on the pixel level. The reconstructed image fidelity of the high-frequency part needs to be improved. For the first time, a perceptual loss is proposed in the field of SPI super-resolution. The perceptual loss describes the perceptual similarity instead of pixel-level similarity. An adversarial loss differentiates the original high-resolution image from the super-resolved image. By combining the two, our results are more natural and more consistent with the perceptual characteristics of human eyes. The superiority of the proposed method over the traditional interpolation methods was visually demonstrated in the experiments. Our simulation shows a peak signal-to-noise ratio of 27.65 dB and structural similarity of 0.8076. Our work shows that GAN-SRSPI is a flexible and effective solution for high-resolution and fast SPI.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.