(Aviation Geophysical Exploration Technology Center, Ministry of Geology and Mineral Resources, Institute of Aviation Geophysical Exploration)
This paper introduces the method and results of displaying, enhancing and interpreting aeromagnetic, aeroradiometric and avionics data in Hami mound survey area of Xinjiang by using digital image processing technology. The image restoration technology studied effectively eliminates the stripe noise caused by the change of atmospheric background in airborne data; This method is also suitable for the pretreatment of avionics data. Through research, three-element image, two-element image, ratio image, divergence image, stereoscopic shadow image, first or second derivative image, local adaptive enhanced image, rose image, gray level segmentation, K-L transform and YIQ—RGB transform are used to extract three aspects of geological information: ① structural features; ② Lithologic mapping; ③ Prospecting anomalies. This work shows the charm of digital image processing technology in displaying and interpreting aviation geophysical data, which has three characteristics: fast, intuitive and easy to synthesize.
I. Introduction
In the field of earth science, digital image processing was originally mainly used for remote sensing. Geophysicists have realized that any spatially varying geophysical data can be displayed and interpreted by digital image processing technology. The parameters used in this work are: airborne radioactivity (total track of potassium, thorium, uranium and TC), aeromagnetism, three-frequency avionics (three frequencies of 520Hz, 2020Hz and 8020Hz, real part and imaginary part) and false color images of satellite photos as reference.
The high-sensitivity integrated aerial survey system used in flight consists of the following equipment: two boxes of NaI crystals with a total volume of 32000cm3, proton precession magnetometer with a sensitivity of 0.5nT and Tridem electromagnetic system. The flying height is 75m, and the measuring scale is 1 ∶ 25000. The variation ranges of potassium content, thorium content and uranium content in this investigation area are 0 ~ 4.6%, 0 ~ 47 ppm and 0 ~ 9. 1ppm respectively. The relative dynamic range of magnetic field is 3540 nT.
Second, the pretreatment of avionics data.
Traditional image processing needs preprocessing first. Due to the instability of atmospheric background, the original aerial data are often accompanied by "banding" phenomenon, and useful information from the earth is often submerged in "banding" noise. In this paper, the image restoration technology is studied. Its principle is shown in figure 1. This method successfully removes the "banded" noise in airborne data.
Fig. 2 shows the original data image of the total trajectory in the upper left corner, the noise image obtained by several moving averages in the upper right corner, the total trajectory image after noise interference is subtracted in the lower left corner, and the final restored image of the total trajectory in the lower right corner.
The image of K, Th and U is actually a regional geochemical map, which is very similar to the satellite image.
Similar to aerial broadcasting, there is a serious "banding" phenomenon in the original avionics data due to the offset value and zero drift of the instrument. Using the above image restoration technology, the avionics image has also been obviously improved. Fig. 3 (color version of fig. 8) shows the comparison before and after the restoration of the 8020Hz real component image. The preprocessing of avionics data also includes image editing, Wallis and median filtering.
Figure 1 Processing Flow of Aerial Image Restoration
Fig. 2 Comparison diagram of air broadcast bus repair
3. Enhancement and interpretation of airborne geophysical data images
Extracting structure tracking information
The directional derivative images of aviation and avionics data contain rich geological structural information, and the stereoscopic shadow image is the most attractive for aeromagnetic extraction of structural features.
Dods et al. (1984) proposed the following formula for calculating the magnetic field stereogram:
Zhang Yujun on new methods of geological exploration.
Where: λ-the included angle between the light source direction and the surface normal; φ —— the height angle of the light source; θ —— Azimuth of light source.
Figure 4 (Figure 8 in color version) is an aeromagnetic color stereoscopic shadow image, in which two variables, the amplitude and gradient of the magnetic field calculated according to the above formula, are displayed at the same time, and the illumination direction is selected as northwest. The color of each pixel represents the total magnetic field, and the brightness of the pixel color changes according to the slope or gradient of the point (Holroyd, 1986).
The bi-directional derivative diagram of magnetic field, color stereoscopic shadow diagram and the image enhanced by local adaptive histogram equalization show the structural characteristics of this area. According to these images, the structural trace map (Figure 5, color plate, Figure 8) was made, and more than 50 structural characteristic lines were determined. The comprehensive interpretation of aeromagnetism, aeroradiation and aeroelectricity is used to study the structural characteristics, and in some cases, supplementary information about the trend of the section can be obtained. The rose diagram shows the frequency statistical distribution of structural features.
Lithologic mapping
RGB-YIQ function is very useful for multi-parameter image enhancement. It converts RGB three bands of a color image into lightness (Y) and chromaticity (I and Q), and approximately, YIQ is equivalent to IHS (lightness, color discrimination and color saturation). Transform and inverse transform according to the following formula:
Zhang Yujun on new methods of geological exploration.
Zhang Yujun on new methods of geological exploration.
In this survey area, we use RGB←→YIQ transformation to achieve two purposes: ① Using the characteristics that the correlation between Y, I and Q components is smaller than that between R, G and B components, we improve the color purity of aerial images through RGB-YIQ-SCALE-RGB processing. ② Enhance the comprehensive parameter table of magnetoelectric amplifier. These two kinds of maps are very useful for lithologic mapping.
A six-band image consists of three aviation elements (K, Th, U) and the amplitudes of three avionics frequencies. Nine lithologic categories are obtained through unsupervised classification (Figure 6): ① ultrabasic rocks; ② copper-nickel ore target area; ③ placer target area; ④ Granite; ⑤ diorite; 6 metamorphic rocks; ⑥ migmatite; ⑧ Quaternary sediments; Pet-name ruby tertiary and quaternary sediments; Attending the tertiary deposit.
The average vectors (k, Th, u) of these categories are listed in the following table 1.
Table 1
Abnormal extraction of copper, nickel and placer
According to the known morphology of low-potassium copper-nickel ore near the fault with local magnetic anomalies, 20 abnormal copper-nickel ore targets were delineated, two of which were proved to be copper-rich nickel mineralization by ground work, and one of them had a copper grade as high as 2%. The average content of potassium in these anomalies is 0.84%.
According to the quaternary sediments with high thorium and uranium content, the abnormal target area of placer is delineated, and the average thorium and uranium contents are 15.4 and 6.09 ppm respectively. The high anomalies of Th and U channels may be caused by zircon and monazite.
Originally published in Beijing (89) International Symposium on Exploration Geophysics, Abstract, 1989.