Translation evaluation based optimized spectrum image data fusion method

Abstract

The present invention discloses a translation evaluation based optimized spectrum image data fusion method, and belongs to the technical field of hyper spectrum image data fusion. The method of the present invention includes the steps: firstly classifying pixel vectors of a data matrix to obtain a class matrix; using a dimensionality reduction and projection matrix with changed principal components of the class matrix to project the data matrix for obtaining a dimensionality reduction data matrix; then assigning the dimensionality reduction data matrix to black and white, red and green, and yellow and blue channels to a coloring matching space for conversion into an sRGB color space to obtain color component data of the R, G, and B channels; comprehensively evaluating serial digital images to obtain an optimal translation amount of each channel through translation mapping of the color component data; and finally using three optimal translation amounts to respectively translate and map the color component data of each channel to obtain a fusion result image. An optimal translation amount fusion method is utilized to obtain a pseudo-color fusion image of high luminance, increases the contrast and color distinction degree of the image, and facilitates rapid identification of a target by human eyes from a complex background.

Description

technical field [0001] The invention belongs to the technical field of hyperspectral image data fusion, and mainly relates to a method for fusing spectral image data cubes into a pseudo-color image, in particular to a fusion of multiple narrow-band spectral image data into a false-color image that is easy to recognize by human eyes method. Background technique [0002] Hyperspectral imaging sensors are based on imaging spectroscopy technology, which simultaneously acquires spatial two-dimensional and spectral information of the detected scene. For any substance, due to the difference in composition, physical and chemical properties, there are differences in its spectral characteristics. This difference can be used to investigate and identify substances, and has been widely used in remote sensing, geology, environment and other fields. [0003] The original spectral data cube acquired by the hyperspectral imaging sensor has a large amount of data, which is difficult to be d...

AI Technical Summary

Problems solved by technology

[0006] The above fusion method converts the principal components of the PCA transformation into the RGB color space, and then compresses the values ​​to the range of 0 to 255 for mapping. The overall brightness of the fused image is dark, and the gray level difference between different target backgrounds of each color channel becomes smaller. The contrast is low, and there is a defect that the display results of similar spectral targets, especially true and false targets, have poor color discrimination and are difficult to distinguish

Images