Abstract
The measurement matrix in compressive imaging controls the crucial feature information for high performance recognition. In this study, a deterministic orthogonal measurement matrix design method using the discrete cosine transform and a compressive feature selection scheme are proposed to implement high-end computational optics for imaging. The selection scheme systematically evaluates the recognition importance for the frequency features, combined with a scaling of the contribution of the various coefficients used to produce a base matrix for the new group of measuring patterns, which ensures the minimal recognition difference for each individual order of frequency filters and combining a relatively complex expression to quickly find the best quantization values. The model parameters are gradually adjusted and eventually converge to the best result through training with a large number of pre-determined samples from the dataset and backpropagating the feature selection loss along with the recognition loss, and the data processing capabilities can be enhanced because the measurement matrix is a priori information for the recognition phase. The systematic ability of the proposed technique was verified through simulations and experiments on two standard datasets: MNIST and CIFAR-10. The results show that the proposed method outperforms state-of-the-art methods in terms of both the model complexity and classification accuracy, which indicates that our study provides a new practical solution for compressive imaging recognition.
Highlights
COMPRESSED sensing (CS) theory [1] is a sampling method that can overcome the limitations of the Nyquist–Shannon sampling theorem, thereby achieving data compression during the sampling while still allowing a recovery of the original information with fewer sampled data and less time
We present a deterministic direct compressivebased measuring method for image recognition and objectoriented imaging based on an orthogonal model derived from a set of discrete cosine transform (DCT) coefficients and trained using optimal feature operators that mimic the visual effect in a convolutional neural networks (CNNs)
The measurement matrix is trained in the parameter optimization module to satisfy the restricted isometry property (RIP) [6] principle in that the measurement results contain the least amount of redundant information when the correlation between measurement patterns is minimal; the measurement patterns are obtained from the DCT basis and are orthogonal to each other
Summary
COMPRESSED sensing (CS) theory [1] is a sampling method that can overcome the limitations of the Nyquist–Shannon sampling theorem, thereby achieving data compression during the sampling while still allowing a recovery of the original information with fewer sampled data and less time. In [20] and [21], the authors designed measurement patterns based on a fast Fourier transform (FFT) and discrete cosine transform (DCT) [22] to extract the compressed frequency features of the images, whereas the authors of [23] used discrete orthogonal Krawtchouk polynomials to measure the image moment information; the measurement patterns of these methods were applied to enhance the image energy (i.e., image information capacity) conventionally in terms of the visual effect, and the clustered features achieved improve the reconstruction efficiency even when discussing the noise immunity and imaging capability [24] in fast reconstruction This shows that, during an image processing cycle, different matrices representing the image information or image energy are varied to optimize a defined function that indicates the reconstructed image quality.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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.
