Limits...
A Real-Time Infrared Ultra-Spectral Signature Classification Method via Spatial Pyramid Matching.

Mei X, Ma Y, Li C, Fan F, Huang J, Ma J - Sensors (Basel) (2015)

Bottom Line: First, we introduce an infrared ultra-spectral signature similarity measure method via SPM, which is the foundation of the matching-based classification method.We calculate the SPM-based similarity between the feature of the spectrum and that of each spectrum of the reference feature library, then take the class index of the corresponding spectrum having the maximum similarity as the final result.Experimental comparisons on two publicly-available datasets demonstrate that the proposed method effectively improves the real-time classification performance and robustness to noise.

View Article: PubMed Central - PubMed

Affiliation: School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan 430074, China. meixiaoguang@hust.edu.cn.

ABSTRACT
The state-of-the-art ultra-spectral sensor technology brings new hope for high precision applications due to its high spectral resolution. However, it also comes with new challenges, such as the high data dimension and noise problems. In this paper, we propose a real-time method for infrared ultra-spectral signature classification via spatial pyramid matching (SPM), which includes two aspects. First, we introduce an infrared ultra-spectral signature similarity measure method via SPM, which is the foundation of the matching-based classification method. Second, we propose the classification method with reference spectral libraries, which utilizes the SPM-based similarity for the real-time infrared ultra-spectral signature classification with robustness performance. Specifically, instead of matching with each spectrum in the spectral library, our method is based on feature matching, which includes a feature library-generating phase. We calculate the SPM-based similarity between the feature of the spectrum and that of each spectrum of the reference feature library, then take the class index of the corresponding spectrum having the maximum similarity as the final result. Experimental comparisons on two publicly-available datasets demonstrate that the proposed method effectively improves the real-time classification performance and robustness to noise.

No MeSH data available.


The spatial pyramid at Level 2 (l = 2). The first row: the two spectra (Spectrum A (left), Spectrum B (right)) are divided into four sub-blocks, respectively. The second row (from the left to the right): the histogram of Block 1, Spec. A, the histogram of Block 2, Spec. A, the histogram of Block 3, Spec. A, the histogram of Block 4, Spec. A, the histogram of Block 1, Spec. B, the histogram of Block 2, Spec. B, the histogram of Block 3, Spec. B, the histogram of Block 4, Spec. B. The third row: the histogram intersection kernels. The fourth row: the summation of the histogram intersection kernels referred to in Equation (2).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541858&req=5

f3-sensors-15-15868: The spatial pyramid at Level 2 (l = 2). The first row: the two spectra (Spectrum A (left), Spectrum B (right)) are divided into four sub-blocks, respectively. The second row (from the left to the right): the histogram of Block 1, Spec. A, the histogram of Block 2, Spec. A, the histogram of Block 3, Spec. A, the histogram of Block 4, Spec. A, the histogram of Block 1, Spec. B, the histogram of Block 2, Spec. B, the histogram of Block 3, Spec. B, the histogram of Block 4, Spec. B. The third row: the histogram intersection kernels. The fourth row: the summation of the histogram intersection kernels referred to in Equation (2).

Mentions: As shown in Figure 3, when l = 2, divide Spectra A and B into four sub-blocks with the same length, respectively. Calculate the histograms of Block 1 to 4 of Spec. A and B. Then, obtain four minimum vectors by comparing the histograms of all sub-blocks of Spec. A with that of Spec. B, respectively. Finally, we have I2 by summing up the four minimum vectors, where I2 is a vector with M dimension.


A Real-Time Infrared Ultra-Spectral Signature Classification Method via Spatial Pyramid Matching.

Mei X, Ma Y, Li C, Fan F, Huang J, Ma J - Sensors (Basel) (2015)

The spatial pyramid at Level 2 (l = 2). The first row: the two spectra (Spectrum A (left), Spectrum B (right)) are divided into four sub-blocks, respectively. The second row (from the left to the right): the histogram of Block 1, Spec. A, the histogram of Block 2, Spec. A, the histogram of Block 3, Spec. A, the histogram of Block 4, Spec. A, the histogram of Block 1, Spec. B, the histogram of Block 2, Spec. B, the histogram of Block 3, Spec. B, the histogram of Block 4, Spec. B. The third row: the histogram intersection kernels. The fourth row: the summation of the histogram intersection kernels referred to in Equation (2).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541858&req=5

f3-sensors-15-15868: The spatial pyramid at Level 2 (l = 2). The first row: the two spectra (Spectrum A (left), Spectrum B (right)) are divided into four sub-blocks, respectively. The second row (from the left to the right): the histogram of Block 1, Spec. A, the histogram of Block 2, Spec. A, the histogram of Block 3, Spec. A, the histogram of Block 4, Spec. A, the histogram of Block 1, Spec. B, the histogram of Block 2, Spec. B, the histogram of Block 3, Spec. B, the histogram of Block 4, Spec. B. The third row: the histogram intersection kernels. The fourth row: the summation of the histogram intersection kernels referred to in Equation (2).
Mentions: As shown in Figure 3, when l = 2, divide Spectra A and B into four sub-blocks with the same length, respectively. Calculate the histograms of Block 1 to 4 of Spec. A and B. Then, obtain four minimum vectors by comparing the histograms of all sub-blocks of Spec. A with that of Spec. B, respectively. Finally, we have I2 by summing up the four minimum vectors, where I2 is a vector with M dimension.

Bottom Line: First, we introduce an infrared ultra-spectral signature similarity measure method via SPM, which is the foundation of the matching-based classification method.We calculate the SPM-based similarity between the feature of the spectrum and that of each spectrum of the reference feature library, then take the class index of the corresponding spectrum having the maximum similarity as the final result.Experimental comparisons on two publicly-available datasets demonstrate that the proposed method effectively improves the real-time classification performance and robustness to noise.

View Article: PubMed Central - PubMed

Affiliation: School of Electronic Information and Communications, Huazhong University of Science and Technology, Wuhan 430074, China. meixiaoguang@hust.edu.cn.

ABSTRACT
The state-of-the-art ultra-spectral sensor technology brings new hope for high precision applications due to its high spectral resolution. However, it also comes with new challenges, such as the high data dimension and noise problems. In this paper, we propose a real-time method for infrared ultra-spectral signature classification via spatial pyramid matching (SPM), which includes two aspects. First, we introduce an infrared ultra-spectral signature similarity measure method via SPM, which is the foundation of the matching-based classification method. Second, we propose the classification method with reference spectral libraries, which utilizes the SPM-based similarity for the real-time infrared ultra-spectral signature classification with robustness performance. Specifically, instead of matching with each spectrum in the spectral library, our method is based on feature matching, which includes a feature library-generating phase. We calculate the SPM-based similarity between the feature of the spectrum and that of each spectrum of the reference feature library, then take the class index of the corresponding spectrum having the maximum similarity as the final result. Experimental comparisons on two publicly-available datasets demonstrate that the proposed method effectively improves the real-time classification performance and robustness to noise.

No MeSH data available.