Limits...
A Multispectral Image Creating Method for a New Airborne Four-Camera System with Different Bandpass Filters.

Li H, Zhang A, Hu S - Sensors (Basel) (2015)

Bottom Line: For this multispectral system, an automatic multispectral data composing method was proposed.For the difficult registration problem between visible band images and near-infrared band images in cases lacking manmade objects, we presented an effective method based on the structural characteristics of the system.Experiments show that our method can acquire high quality multispectral images and the band-to-band alignment error of the composed multiple spectral images is less than 2.5 pixels.

View Article: PubMed Central - PubMed

Affiliation: Key Laboratory of 3D Information Acquisition and Application of Ministry, Capital Normal University, Beijing 100048, China. lihanlun@126.com.

ABSTRACT
This paper describes an airborne high resolution four-camera multispectral system which mainly consists of four identical monochrome cameras equipped with four interchangeable bandpass filters. For this multispectral system, an automatic multispectral data composing method was proposed. The homography registration model was chosen, and the scale-invariant feature transform (SIFT) and random sample consensus (RANSAC) were used to generate matching points. For the difficult registration problem between visible band images and near-infrared band images in cases lacking manmade objects, we presented an effective method based on the structural characteristics of the system. Experiments show that our method can acquire high quality multispectral images and the band-to-band alignment error of the composed multiple spectral images is less than 2.5 pixels.

No MeSH data available.


The true-color composite (red, green, blue); (a) the unregistered multispectral image; (b) the enlarged partial region of (a); (c) the registered multispectral image; (d) the enlarged partial region of (c).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541943&req=5

sensors-15-17453-f017: The true-color composite (red, green, blue); (a) the unregistered multispectral image; (b) the enlarged partial region of (a); (c) the registered multispectral image; (d) the enlarged partial region of (c).

Mentions: Compared with Figure 4 and Figure 6, Figure 8 contains no man-made objects. The four methods are used respectively to compose one four-band image. Figure 17a shows the true-color composite of the unregistered multispectral image, and Figure 17b is its enlarged partial region. The dislocation is obvious. Figure 17c,d show the true-color composite of the registered multispectral image. There is no dislocation between different bands. Because there are no artificial objects in these images, there is a great difference between the visible band SIFT features and infrared band SIFT features, and the correct rate of initial matches is lower than 50%, so the first three methods cannot get a correct infrared band in the four-band image, as shown in the first three severely distorted images of Figure 18. Compared with the other three methods, the fourth method uses the rejecting false matches method mentioned in the Section 3 to remove most false matches firstly, and then uses the RANSAC; therefore it can get a correct infrared band, as shown in the fourth image of Figure 18. So, the first three methods cannot get the CIR composite correctly, as shown in the first three images of Figure 19, but the fourth method can, as shown in the fourth image of Figure 19. Table 4 shows that the errors of the third method and the fourth method are the same at the column G-B and G-R, and a little better than the first and second method. At the G-IR column, the first three methods cannot get correct results, but the fourth method can get correct results with a very low error, about 2.5 pixels.


A Multispectral Image Creating Method for a New Airborne Four-Camera System with Different Bandpass Filters.

Li H, Zhang A, Hu S - Sensors (Basel) (2015)

The true-color composite (red, green, blue); (a) the unregistered multispectral image; (b) the enlarged partial region of (a); (c) the registered multispectral image; (d) the enlarged partial region of (c).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541943&req=5

sensors-15-17453-f017: The true-color composite (red, green, blue); (a) the unregistered multispectral image; (b) the enlarged partial region of (a); (c) the registered multispectral image; (d) the enlarged partial region of (c).
Mentions: Compared with Figure 4 and Figure 6, Figure 8 contains no man-made objects. The four methods are used respectively to compose one four-band image. Figure 17a shows the true-color composite of the unregistered multispectral image, and Figure 17b is its enlarged partial region. The dislocation is obvious. Figure 17c,d show the true-color composite of the registered multispectral image. There is no dislocation between different bands. Because there are no artificial objects in these images, there is a great difference between the visible band SIFT features and infrared band SIFT features, and the correct rate of initial matches is lower than 50%, so the first three methods cannot get a correct infrared band in the four-band image, as shown in the first three severely distorted images of Figure 18. Compared with the other three methods, the fourth method uses the rejecting false matches method mentioned in the Section 3 to remove most false matches firstly, and then uses the RANSAC; therefore it can get a correct infrared band, as shown in the fourth image of Figure 18. So, the first three methods cannot get the CIR composite correctly, as shown in the first three images of Figure 19, but the fourth method can, as shown in the fourth image of Figure 19. Table 4 shows that the errors of the third method and the fourth method are the same at the column G-B and G-R, and a little better than the first and second method. At the G-IR column, the first three methods cannot get correct results, but the fourth method can get correct results with a very low error, about 2.5 pixels.

Bottom Line: For this multispectral system, an automatic multispectral data composing method was proposed.For the difficult registration problem between visible band images and near-infrared band images in cases lacking manmade objects, we presented an effective method based on the structural characteristics of the system.Experiments show that our method can acquire high quality multispectral images and the band-to-band alignment error of the composed multiple spectral images is less than 2.5 pixels.

View Article: PubMed Central - PubMed

Affiliation: Key Laboratory of 3D Information Acquisition and Application of Ministry, Capital Normal University, Beijing 100048, China. lihanlun@126.com.

ABSTRACT
This paper describes an airborne high resolution four-camera multispectral system which mainly consists of four identical monochrome cameras equipped with four interchangeable bandpass filters. For this multispectral system, an automatic multispectral data composing method was proposed. The homography registration model was chosen, and the scale-invariant feature transform (SIFT) and random sample consensus (RANSAC) were used to generate matching points. For the difficult registration problem between visible band images and near-infrared band images in cases lacking manmade objects, we presented an effective method based on the structural characteristics of the system. Experiments show that our method can acquire high quality multispectral images and the band-to-band alignment error of the composed multiple spectral images is less than 2.5 pixels.

No MeSH data available.