Method for monitoring the stability of an excavation face using radar interferometry

Abstract

To remove an effect induced by a mobile object in an interference image.SOLUTION: A method includes periodical acquisition in nc pieces of cycles of images Si of a scenario, each of the images Si includes np pieces of pixels Pij having a spatial coordinate defined with respect to a predetermined reference system. Here, Pij is a j-th pixel acquired in an i-th cycle. The method includes repetition of steps of calculation of a coherence value Yij arranged so as to express quality of phase information provided by the j-th pixel in the i-th cycle, comparison with a predetermined minimum coherence value Ymin, and interference calculation of a displacement value dij of the pixel Pij, in the case of Yij<Ymin, the displacement value dij is 0, calculation of a cumulative displacement value Dij as the total sum of the displacement value dij in an N cycle preceding the i-th cycle, in which the cumulative displacement value Dij is according to Expression.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] Specification The present invention relates to the field of monitoring the stability of an excavation face by synthetic aperture radar (SAR) interferometry.

[0002] In particular, the present invention relates to a method for removing the effects induced by moving objects in an interference image.

[0003] Description of the Related Art As is well known, at present, continuous monitoring devices based on laser technology are used to monitor the stability of an excavation face. The purpose of these monitoring systems is to provide information regarding displacements and deformations that are precursors of the possibility of collapse of the excavation face in order to enable the work to be stopped by securing personnel and machinery.

[0004] In most cases, with these laser systems, in order to detect any movement with an accuracy of approximately 1 mm, it is necessary to install an optical reflector fixed corresponding to the area to be monitored and measure the position periodically.

[0005] However, the installation of reflectors on the excavation face is not very simple because the continuous processing operations and the continuous evolution of the face itself necessitate the periodic repositioning of the reflectors. Furthermore, laser-based systems are prone to deterioration of measurements in the presence of dust or steam resulting from the excavation operation.

[0006] The evolution of this measurement system is represented by a system called OBM described in "Advanced observation techniques for sophisticated shallow tunnel projects - experience gained using innovative monitoring methods at the Lainzer Tunnel LT31" (Moritz Bernd et al., "Geomechanik a Tunnelbau: Geomechanik a Tunnelbau" 1.5, 2008, pp.466 - 476), which uses a total station without applying a reflector to the excavation face. Also, a fan that always blows towards the total station is used to keep the lens of the instrument free of dust.

[0007] However, the accuracy of this system is 2 - 4 mm, which is significantly lower than that of a laser system using an optical reflector, and there is a risk of reduced monitoring effectiveness.

[0008] An alternative solution is the use of an interferometric radar, which is a technology widely used for the important monitoring of landslides in mining situations. Radar interferometry is more effective in detecting precursors of the possibility of excavation face collapse because it provides displacement measurements with sub - millimeter accuracy. At the same time, this technology is less affected by environmental disturbances such as dust and vapor.

[0009] However, this technology has an important drawback when applied to the excavation face, that is, the presence of disturbances caused by machines operating near the excavation face degrades the quality of the radar data, and as a result, it becomes ineffective for monitoring movement.

Summary of the Invention

[0010] Therefore, an object of the present invention is to provide a method for monitoring the stability of an excavation face by SAR interferometry that enables the detection of displacements with sub - millimeter accuracy even in the presence of moving machines.

[0011] Another object of the present invention is also to provide a method that filters interference data, preserves only a part of the information that can be used for displacement measurement, and enables avoidance of generating false alarms due to disturbances generated by machines.

[0012] It is still an object of the present invention to provide such a method that enables providing a three-dimensional representation of displacement data by combining laser measurement and radar interference measurement.

[0013] These and other objects are achieved by a method for filtering interferometric radar acquisitions, the method comprising: · A pre-arrangement of a radar system for performing acquisition of an image of a scenario by SAR interferometry, the radar system comprising: · At least one radar sensor arranged to emit and receive radar signals, and · A control unit configured to analyze the signals received by the radar sensor by means of interference techniques, and · A screen arranged to display the image of the scenario to the user, the pre-arrangement, and · Periodic acquisition by the radar system of an image S i in a number n c of cycles, where i = 1, 2,..., n c and each image S i has spatial coordinates defined with respect to a given reference system, the number n p of pixels P ij where P ij is the j-th pixel of the i-th image acquired in the i-th cycle, where j = 1, 2,..., n p and the periodic acquisition, including steps. In particular, for each i-th cycle and for each pixel P ij a coherence value Y arranged to represent the quality of the phase information supplied by the j-th pixel in the i-th cycleij Calculation of · Coherence value Y ij with a predetermined minimum coherence value Y min and comparison with · The pixel P ij displacement value d ij interference calculation, where Y ij < Y min in which case, the displacement value d ij is zero, interference calculation, and · The cumulative displacement value D as the sum of the displacement values d in N cycles preceding the i-th cycle ij is calculated, and the calculation of the cumulative displacement value D ij is calculated according to the formula, and an iteration of steps of the calculation is provided ij is provided

Equation

[0014] In particular, the coherence value Y ij is arranged to represent the degree of correlation of the phase information provided by the j-th pixel in the i-th cycle using the phase information provided by the same pixel in the previous cycle.

[0015] Advantageously, each pixel P ij is · A mask coefficient M arranged to take a value M ij = NaN or a value M ij = 1, where when M ij = NaN, the pixel P ij is not shown on the screen, and when M ij = 1, the pixel P ij is shown on the screen, the mask coefficient Mij, and ij · When Yij < Y a value m min = 0, or when Yij ≧ Y ij a value m min = 1 is arranged to take a coherence coefficient M ij is associated with ij and

[0016] Specifically, in each i-th cycle and for each pixel P ij with respect to · Evaluation of the coherence coefficient m in the N cycles preceding the i-th cycle ij and · Value threshold excess value f as the sum of the coherence coefficient m in the N cycles preceding the i-th cycle excluding the number of cycles N ij Calculation, where the value threshold excess value ij is calculated according to the formula ij and

Equation

Equation

[0017] Advantageously · Calculation of the power value W for each pixel P i of the predetermined image S ij and ij · Comparison of the power value W for each pixel P ij with a predetermined minimum power value W ij and min · Assignment of the static mask coefficient · W ij <W min to the pixel P

Equation

Number

[0018] More specifically, the following conditions are valid.

Number

Number

Number

Number

[0019] Advantageously,[[]] · Laser scanning by obtaining a 3D map of the scenario, and · Each cumulative displacement value D on the 3D map of the scenario obtained by the step of laser scanning ij The step of projection of and is provided.

[0020] In particular, the cumulative displacement value D ij When it exceeds a predetermined threshold D max The radar system is adapted to issue an alarm.

[0021] Advantageously, the mask coefficient M ij = NaN, the step of calculating the pixels P of the number q ij is provided, and when the ratio q / n p exceeds a predetermined threshold, the radar system is adapted to issue an alarm.

[0022] Advantageously, in each i-th cycle and for each pixel P ij Regarding · Equation

Number

Number

[0023] In particular, the displacement value d ij is calculated according to the formula

Number

[0024] Advantageously, the coherence value Y ij is a spatial coherence value, which is calculated by the formula

Number

Number

Number

Brief Description of the Drawings

[0025] Further features and / or advantages of the present invention will become more apparent from the following description of its exemplary, but non-limiting, embodiments with reference to the accompanying drawings.

[0026]

Figure 1A

Figure 1B

Figure 2A

Figure 2B

Figure 3

Figure 4

Figure 5

MODE FOR CARRYING OUT THE INVENTION

[0027] Referring to FIGS. 1 and 4, in a possible embodiment of the present invention, a method for filtering interference radar acquisitions includes a first step of pre-positioning a radar system arranged to provide acquisition of an image of a scenario by SAR interferometry. In particular, the radar system comprises at least one radar sensor 110 arranged to emit and receive radar signals, and a screen 120 arranged to display an image of the scenario. The radar system also comprises a control unit 115 configured to analyze the signals received by the radar sensor 110 by means of interference techniques and transmit the image to the screen 120.

[0028] Next, the method is such that the radar system acquires an image S of the scenario ifor i = 1, 2, …, n c for the number n c provides a step of periodic acquisition over a cycle. Each image S i has spatial coordinates defined with respect to a predetermined reference system, for the number n p of pixels P ij where the j-th pixel of the i-th image acquired in the i-th cycle, and j = 1, 2, …, n p is.

[0029] This method then provides, for each i-th cycle, the process of image acquisition by the control unit by performing the following steps repeatedly for each pixel P ij .

[0030] There is a first step of calculating an interferogram of the relative pixel P ij according to the formula,

Number

Number

[0031] There is then a step of calculating a coherence value Y ij arranged to represent the quality of the phase information supplied by the j-th pixel in the i-th cycle. In particular, the coherence value Y ij is arranged to represent the degree of correlation of the phase information provided by the j-th pixel in the i-th cycle using the phase information provided by the same pixel in the previous cycle.

[0032] In one embodiment of the present invention, the coherence value may be a spatial coherence value calculated taking into account the i-th cycle and the previous cycle by the formula,

Number

Number

[0033] In the subsequent step, the calculated spatial coherence value Y ij is compared with a predetermined minimum spatial coherence value Y min . Each pixel P ij is also associated with a coherence coefficient m ij that can take the value m ij = 0 or the value m ij = 1.

[0034] Y ij < Y min , the coherence coefficient value m ij = 0 is assigned to the above pixel P ij , while in the case of Yij ≧ Y min , the coherence coefficient value m ij = 1 is assigned to the above pixel P ij .

[0035] Next, there is a step of interference calculation of the displacement value d ij of each pixel P ij according to the formula,

Number

[0036] Finally, in the N cycles preceding the i-th cycle, the displacement value dij The cumulative displacement value D as a sum, that is, according to the following formula ij There is a step of calculation.

Number

[0037] Therefore, the method claimed in the present invention assigns a unitary value or a null value to the coherence coefficient m ij associated with each pixel P ij to exclude unreliable pixels from the calculation of the cumulative displacement D because they do not have sufficient coherence, that is, the movement of objects with low coherence within the scenario results in low coherence. ij This makes it possible to exclude them from the calculation of the cumulative displacement D.

[0038] For example, the machine 200 shown in FIGS. 1 and 2A, when turned on, vibrates and moves very fast with respect to the dimensions and scanning time of the radar pixels, resulting in objects with low coherence. On the other hand, an excavation surface, such as a tunnel, is a very coherent object because it moves slowly and consistently. In FIG. 2B, it is possible to confirm that the zone A corresponding to the presence of the machine 200 in FIG. 2A has pixels with low coherence (darker color) compared to the pixels in the surrounding zones.

[0039] As also referenced in FIG. 5, in a modified embodiment of the method according to the present invention, when M ij = NaN and is shown instead of Mij = 1, the value M ij = NaN or a mask coefficient M arranged to take the value Mij = 1 ij is also provided for the association with each pixel P ij of the mask coefficient M.

[0040] Therefore, such mask coefficient values M ijIt is possible to determine the non-display of a pixel from the screen, for example, when this pixel has insufficient power (static mask) or when the coherence coefficient value is low for too long compared to a predetermined threshold value (dynamic mask).

[0041] In particular, in this variant embodiment, the method is based on a predetermined image S i for each pixel P ij of the power value W ij implemented by a first step of calculation of the static mask filter, this power value W ij with a predetermined minimum power value W win and a comparison step, and, when W ij <W min assigns the static mask coefficient

Number

[0042] This filter in the display of the pixels on the screen 120 is defined as a "static" mask filter because the aforementioned steps are not repeated in each cycle but are only executed when required on a specific image. For example, the static mask filter can be implemented on both the image acquired at the beginning of the scenario and the next image, and, if necessary, affects all subsequent images as long as no new calculation of the signal power is performed.

[0043] This calculation is performed using an image acquired with no machine in order to measure the signal strength at the excavation face. The static mask filter makes it impossible to display pixels with too low a signal strength, which may result in false detections. That is, it reports displacements that do not actually occur but appear to occur due to high background noise.

[0044] Still referring to FIG. 5, in this variant embodiment, the method also, in each i-th cycle and for the coherence coefficient value m in the N cycles preceding the i-th cycle ijFor each pixel P in the first step of the evaluation ij includes a dynamically applied mask filter. Then, the coherence coefficient m in the N cycles preceding the i-th cycle, divided by the number of cycles N, i.e., according to the formula ij The sum of the threshold exceedance values f ij The calculation step follows

Number

[0045] Therefore, the threshold exceedance value f ij The comparison step with the predetermined minimum threshold exceedance value f min And, for f ij < f min In the case of, the dynamic mask coefficient

Number

[0046] This filter in the display of pixels on the screen 120 is defined as a "dynamic" mask filter because the above steps are repeated in each cycle, preventing the display of pixels that are considered untrustworthy because they have insufficient spatial coherence over several consecutive cycles

[0047] This may exclude pixels from the image that have the potential to cause detection omissions. That is, in reality, although movement is occurring, it reports that no movement is occurring

[0048] As schematized in FIG. 5, the condition

Number

Number

[0049] Thus, the static and dynamic mask filters enable only truly reliable pixels to be displayed on the screen 120, providing the operator with an immediate understanding of the reliability of the image. In FIG. 3, a screen representation of pixel displacement values projected onto a three-dimensional map obtained by laser scanning of the scenario can be seen. In particular, zone A corresponds to a zone with low pixel coherence and thus is not displayed on the map, and zone B shows pixels with large displacements. Since the pixels in zone B have high coherence, false detections can be excluded and the displacement measurements can be considered reliable.

[0050] Some of the exemplary specific embodiments in the foregoing description fully clarify the invention from a conceptual point of view, so that others can, by applying current knowledge, modify and / or adapt the specific exemplary embodiments in various applications without further research and without departing from the invention. Thus, such adaptations and modifications are meant to be considered equivalent to the specific embodiments. The means and materials for realizing each function described herein may, for this reason, have different natures without departing from the field of the invention. It should be understood that the syntax or terminology used herein is for the purpose of description and not of limitation.

Claims

1. A method for filtering interferometric radar acquisitions, comprising: A pre - arrangement of a radar system for performing acquisition of an image of a scenario by SAR interferometry, said radar system comprising: At least one radar sensor (110) arranged to emit and receive radar signals; A control unit (115) configured to analyze said signals received by said radar sensor (110) by means of interference techniques; A screen (120) arranged to display said image of said scenario to a user; Said pre - arrangement, comprising: The radar system acquires the image S of the scenario ｉ at a cycle of number n ｃ in a periodic acquisition, where i = 1, 2, …, n ｃ and each image S ｉ has spatial coordinates defined with respect to a predetermined reference system and includes number n ｐ pixels P ｉｊ where P ｉｊ is the j-th pixel of the i-th image acquired in the i-th cycle, where j = 1, 2, …, n ｐ The periodic acquisition and Steps of: In each i-th cycle and for each pixel P ｉｊ with respect to Coherence value Y arranged to represent the quality of the phase information supplied by the j-th pixel in the i-th cycle ｉｊ the calculation of, the coherence value Y ｉｊ with a predetermined minimum coherence value Y ｍｉｎ comparison therewith, the pixel P ｉｊ has a displacement value d ｉｊ is an interference calculation, where Y ｉｊ < Y ｍｉｎ in the case of, the displacement value d ｉｊ is zero, the interference calculation, and The cumulative displacement value D as the sum of the displacement values d in the N cycles preceding the i-th cycle ｉｊ is calculated, and the cumulative displacement value D ｉｊ is such that the formula ｉｊ holds 【Number 1】 Calculations calculated according to: Iterations of steps of: being provided, characterized in that Each pixel Pij is associated with a coherence coefficient mij arranged to take the value mij = 0 when Yij < Ymin or the value mij = 1 when Yij ≥ Ymin; Said interference calculation comprising: An equation 【Number 2】 A calculation of the interferogram Iij of said pixel Pij according to: where sij is a complex number representing the focused radar data of the j - th pixel of the i - th image; [Number 3] A step of calculation where is the complex conjugate of sij; An interference calculation of said displacement value dij as a function of arg[Iij], where arg[Iij] is the argument of the complex number Iij; Said displacement value dij being calculated according to the equation [Number 4] According to: where λ is the wavelength of the signal emitted by said radar system, a step of interference calculation and Including: A method.

2. Each pixel P ｉｊ is Value M ｉｊ = NaN or value M ｉｊ Mask coefficient M arranged to take = 1 ｉｊ where M ｉｊ when = NaN, the pixel P ｉｊ is not shown on the screen (120), and when M ｉｊ = 1, the pixel P ｉｊ is shown on the screen (120), mask coefficient Mij The method for filtering interferometric radar acquisitions according to Claim 1, associated with:

3. In each i-th cycle and for each pixel P ｉｊ with respect to The evaluation of the value of the coherence coefficient m in the N cycles preceding the i-th cycle ｉｊ and The value of the excess fraction value f as the sum of the coherence coefficients m in the N cycles preceding the i-th cycle, excluding the number of cycles N ｉｊ is calculated, and the excess fraction value f ｉｊ is given by the formula ｉｊ where 【Number 5】 Calculations calculated according to: The threshold exceedance fractional value f ｉｊ with a predetermined minimum exceedance threshold fractional value f ｍｉｎ for comparison, and f ｉｊ <f ｍｉｎ In the case of, the dynamic mask coefficient 【Number 6】 said pixel P ｉｊ The method according to claim 2, further comprising a step of assigning to and, for filtering the interference radar acquisition.

4. The specified image S ｉ for each pixel P ｉｊ of the power value W ｉｊ calculation and, For each pixel P ｉｊ with respect to the power value W ｉｊ and a predetermined minimum power value W ｍｉｎ comparison therewith W ｉｊ <W ｍｉｎ in the case of, static mask coefficient 【Number 7】 said pixel P ｉｊ assignment to, and W ｉｊ ≥W ｍｉｎ in the case of, static mask coefficient 【Number 8】 said pixel P ｉｊ A method for filtering the interference radar acquisition according to claim 2, wherein an assignment to and a step of are provided.

5. 【Number 9】 【Number 10】 【Number 11】 【Number 12】 The method for filtering interferometric radar acquisitions according to Claims 3 and 4, being:

6. Laser scanning by obtaining a three - dimensional map of said scenario; Each cumulative displacement value D on the three-dimensional map of the scenario obtained by the step of laser scanning ｉｊ and the step of are provided, a method for filtering the interference radar acquisition according to claim 1.

7. Cumulative displacement value D ｉｊ When it exceeds a predetermined threshold value D ｍａｘ The method for filtering the interference radar acquisition according to claim 1, wherein the radar system is adapted to issue an alarm.

8. Mask coefficient M ｉｊ = Pixel P having NaN ｉｊ A step of calculating the number q is provided, and when the ratio q / n ｐ exceeds a predetermined threshold value, the radar system is adapted to issue an alarm. The method for filtering interference radar acquisition according to claim 2

9. the coherence value Y ｉｊ is a spatial coherence value, and it 【Number 13】 Calculated by the formula of: where 【Number 14】 Is a spatial media mathematical operator operating in the vicinity of the j - th pixel; Here, s ｉｊ is the complex number representing the focused radar data of the j-th pixel of the i-th image, 【Number 15】 is the complex conjugate of s ｉｊ The method according to claim 1 for filtering the interference radar acquisition, wherein [the value] is the complex conjugate of s

Images