Abstract
When measurement rates grow, most Compressive Sensing (CS) methods suffer from an increase in overheads of transmission and storage of CS measurements, while reconstruction quality degrades appreciably when measurement rates reduce. To solve these problems in real scenarios such as large-scale distributed surveillance systems, we propose a low-cost image CS approach called MRCS for object detection. It predicts key objects using the proposed MYOLO3 detector, and then samples the regions of the key objects as well as other regions using multiple measurement rates to reduce the size of sampled CS measurements. It also stores and transmits half-precision CS measurements to further reduce the required transmission bandwidth and storage space. Comprehensive evaluations demonstrate that MYOLO3 is a smaller and improved object detector for resource-limited hardware devices such as surveillance cameras and aerial drones. They also suggest that MRCS significantly reduces the required transmission bandwidth and storage space by declining the size of CS measurements, e.g., mean Compression Ratios (mCR) achieves 1.43–22.92 on the VOC-pbc dataset. Notably, MRCS further reduces the size of CS measurements by half-precision representations. Subsequently, the required transmission bandwidth and storage space are reduced by one half as compared to the counterparts represented with single-precision floats. Moreover, it also substantially enhances the usability of object detection on reconstructed images with half-precision CS measurements and multiple measurement rates as compared to its counterpart, using a single low measurement rate.
Highlights
Challenges and MotivationsHigh-resolution cameras and drones are increasingly used for capturing images in large-scale distributed surveillance systems
To reduce the required transmission bandwidth and storage space for Compressive Sensing (CS) measurements, we propose a CS approach with multiple Measurement Rates (MRs) for sampling natural images known as MRCS
The performance of CS sampling and reconstruction with the proposed MRCS is illustrated in Table 3, where AP0.5 denotes that the accuracy for a class of objects predicted with MYOLO3 on the VOC-revise test dataset when the value of Intersection over Union (IoU) is 0.5
Summary
High-resolution cameras and drones are increasingly used for capturing images in large-scale distributed surveillance systems. Massive amounts of images and videos collected over large-scale surveillance networks make the timely transmission of all captured data to remote servers as well as permanent storage an impractical one. For large-scale distributed surveillance systems, it is a great challenge to reduce the required transmission bandwidth and storage space while preserving the usability of sampled images/videos. The CS method is especially fit for resource-constrained image capturing devices such as aerial drones and robots since it has the potential to significantly reduce the cost of image/video transmission and storage. The size of measurement data increases rapidly as measurement rate grows It is very challenging for CS sampling to use as few measurements as possible while recovering images with adequate local features that users are interested in. CS measurements sampled with multiple MRs, which is employed as a post-acquisition step on high-performance servers [8]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
