Multi-resolution reconstruction algorithm for compressive single pixel spectral imaging

García, Hans; Correa, Claudia V.; Villarreal, Ofelia P.; Pinilla, Samuel; Argüello, Henry

doi:10.23919/eusipco.2017.8081251

Cited by 8 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…UAV system for mapping a cocoa field PC5:Implementation of drones for the identification of pests and diseases. Multi-resolution reconstruction algorithm [44], Hierarchical Feature Extraction Model [45], Convolutional Neural Network [46], Selective multi-convolutional region (SMCR) [46]. [11], [27], [31], [32], [33], [34] PC2 (Pre-Processing Model)…”

Section: Based On Tablementioning

confidence: 99%

A Review of Image Processing Techniques for Pest and Disease Early Identification Systems on Modern Cocoa Plantation

Basri,

Indrabayu,

Achmad

et al. 2023

IJCDS

View full text Add to dashboard Cite

This study presents a potential approach to applying image processing to the cocoa farming industry. Industry 4.0 is changing how technology is applied in cocoa cultivation. Research shows that pests and diseases are major factors that can reduce productivity, especially in the upstream and downstream industries. This study aims to conduct a literature review of several related studies, including previous research. Based on relevant references related to previous research, it is found that the case of disease attack in cocoa is most often studied. This study shows that the color analysis model is more inclined to the case of disease image patterns. The performance pattern of the existing approaches, the average accuracy is 82.85%. More challenges in image processing and efforts are made to improve accuracy but with lower processing complexity. This study also recommends a model that supports the implementation of image processing with machine learning technology in the upstream cocoa cultivation industry. The study results show that the proposed modeling framework can be the following reference for research development towards artificial intelligence technology to support Industry 4.0, particularly in cocoa cultivation.

show abstract

Section: Based On Tablementioning

confidence: 99%

A Review of Image Processing Techniques for Pest and Disease Early Identification Systems on Modern Cocoa Plantation

Basri,

Indrabayu,

Achmad

et al. 2023

IJCDS

View full text Add to dashboard Cite

show abstract

“…In this section, the Hadamard matrix is used to design the sensing matrix since its rows are mutually orthogonal; this property is desired in compressive sensing [12,23]. This property allows a fast reconstruction approach, due to the transpose normally used in the GPSR algorithm is reduced to only one matrix product [18,24]. Thus, Equation.29 isC…”

Section: Design Of the Sampling Matrix Based On Hadamard Matricesmentioning

confidence: 99%

Single Pixel Compressive Spectral Polarization Imaging using a Movable Micro-Polarizer Array

Bacca-Quintero

Vargas-García

Molina-Velasco

et al. 2018

Rev. Fac. Ing. Univ. Antioquia

View full text Add to dashboard Cite

The acquisition of spectral polarization images is a method that obtains polarized, spectral and spatial information of a scene. Traditional acquisition methods use dynamic elements that capture all the information of a scene, by scanning the areas of interest, which result in large amounts of data proportional to the desired image resolution. Hence, in this work, the compression of spectral polarization images using a single pixel architecture, that uses a micro-polarizer array aligned with a binary coded aperture is proposed. The micro-polarizer is moved horizontally in each shot, so that diverse types of codifications from the scene are obtained. The proposed architecture allows several compressive 2D projections with spatial, spectral and polarization coding to be obtained. Stokes parameter images at several wavelengths are reconstructed. This architecture reduces the total number of measurements needed to obtain spectral polarization images compared to traditional acquisition methods. The experiments validate the quality of the architecture obtaining 43.19 dB, 37.49 dB and 30.41 dB of the peak signal-to-noise ratio for the first three Stokes parameters respectively. RESUMEN: La adquisición de imágenes espectro polarizadas es un método que obtiene información espacial, espectral y de polarización de una escena. Los métodos tradicionales de adquisición utilizan elementos dinámicos que capturan la totalidad de la información de la escena, escaneando las áreas de interés. Esto resulta en grandes cantidades de datos proporcionales a la resolución de imagen deseada. Por esta razón, en este trabajo se propone la compresión de imágenes espectro polarizadas usando una arquitectura óptica de único píxel, que usa un arreglo de micro-polarizadores alineados con una apertura codificada binaria. El arreglo de micro-polarizadores es movido horizontalmente en cada captura, permitiendo lograr diversos tipos de codificaciones de la escena. La arquitectura propuesta permite obtener varias proyecciones 2-D comprimidas con codificación espacial, espectral y de polarización, para luego reconstruir los Parámetros de Stokes a varias longitudes de onda. Esta arquitectura reduce el numero total de medidas necesarias para obtener la información espectral y de polarización de las imágenes comparadas con las arquitecturas tradicionales. Los experimentos validad la calidad de la arquitectura propuesta obteniendo 43.19 dB, 37.49 dB y 30.41 dB de la proporción máxima de señal a ruido para los tres primeros de Stokes respectivamente.

show abstract

“…In this work, a multi-resolution (MR) reconstruction model is proposed such that the computational complexity of the inverse problem is reduced by employing super-pixels, which in turn reduces the number of unknown values to recover in the inverse problem. Given that most scenes exhibit areas with similar spectral signatures, several pixels can be grouped into super-pixels without losing information or inducing considerable errors [12]- [15]. For instance, super-pixels have been previously used for gray-scale image segmentation [16]- [18].…”

Section: Introductionmentioning

confidence: 99%

SLIC super-pixels for multi-resolution compressive spectral imaging reconstruction

García¹,

Correa²,

Fuentes³

2018

Opt. Pura Apl.

View full text Add to dashboard Cite

Spectral imaging (SI) is widely used in different applications involving material identification since it contains both spatial ( , ) and spectral information ( ). However, traditional SI scanning methods involve massive amounts of data, which increase the cost of storing and processing. Compressive sensing (CS) theory has been applied in SI, such that the underlying data cube can be recovered from a reduced number of measures. Reconstructions are obtained by 2 − 1 norm-based algorithms whose computational complexity grows in proportion to the number of unknowns. In this paper, a multiresolution reconstruction model based on the simple linear iterative clustering (SLIC) is proposed to reduce the number of unknown values to recover. Simulation results show that the proposed method is up to 86% faster than the full-resolution reconstructions. Additionally, MR approach obtains more accurate reconstructions with improvements of up to 12dB of PSNR.

show abstract

Multi-resolution reconstruction algorithm for compressive single pixel spectral imaging

Cited by 8 publications

References 15 publications

A Review of Image Processing Techniques for Pest and Disease Early Identification Systems on Modern Cocoa Plantation

A Review of Image Processing Techniques for Pest and Disease Early Identification Systems on Modern Cocoa Plantation

Single Pixel Compressive Spectral Polarization Imaging using a Movable Micro-Polarizer Array

SLIC super-pixels for multi-resolution compressive spectral imaging reconstruction

Contact Info

Product

Resources

About