A coded-aperture technique and laser radar-based radioactive source quantification imaging method

Abstract

The application discloses a kind of based on coding aperture technology and laser radar's radioactive source quantification imaging method, its steps include: 1) using coding aperture gamma camera obtains the coding projection and energy spectrum of radioactive source in target area, and decoding calculation is carried out to coding projection to obtain coding reconstruction image O;2) obtaining the radiation hot spot in image O, generates radiation hot spot image;3) according to energy spectrum, the full energy peak E of radioactive source is identified, and the dose rate at the position where camera is calculated;4) by laser radar, the three-dimensional point cloud information of the target area is acquired, and the optical video information of target area is acquired by optical camera;Then according to azimuth, the pixels in radiation hot spot image, optical video information, three-dimensional point cloud information are matched and fused;5) find the three-dimensional point cloud pixel corresponding to radiation hot spot in fusion map;According to the three-dimensional point cloud pixel corresponding to radiation hot spot, the distance between radiation source and camera is obtained;6) inversion calculation obtains the activity value of radioactive source.

Description

Technical Field

[0001] The present invention is applied to radiation imaging fields involving gamma-ray sources, such as radiation protection monitoring, nuclear emergency response, and radioactive source search in nuclear facilities. It can obtain the distribution information of imaging radiation hotspots and provide the activity values ​​of imaging radiation hotspots at a long distance without contacting the radioactive source, thereby optimizing and improving radioactive source imaging technology, and particularly relates to a radioactive source quantitative imaging method based on coded aperture technology and lidar. Background Art

[0002] In the monitoring and investigation of radiation sources, nuclear radiation imaging technology has become an advanced radiation measurement technique, capable of rapidly measuring the overall radiation environment and locating key radiation hotspots. Existing nuclear radiation imaging methods can produce two-dimensional pseudo-color images of the radiation source distribution. By fusing these images with optical video footage, the azimuth of the radiation source can be presented as hotspots. However, existing imaging techniques can only provide relative intensities of radiation hotspots. For example, in source surveys of nuclear facilities, when multiple radiation hotspots appear in the field of view, their true strength cannot be compared due to their varying distances. To obtain the intensity value (activity or surface dose rate) of the radiation source during imaging, it is necessary to obtain the distance value to the hotspot from the source, and then inversely calculate its intensity. Multi-messenger fusion technology is also currently being applied in the field of radiation imaging. When imaging a radiation source, in addition to obtaining radiation information, an optical video camera can be added to the system to obtain optical information, enabling more intuitive visualization of the radiation source. With advances in lidar technology, application cases in which lidar modules are installed in imaging systems have begun to emerge in recent years. For example, a LiDAR rangefinder and a Compton (or coded aperture) imaging camera can be integrated into a robotic platform to perform radar scanning modeling of a 3D scene by roaming through a radiation environment. This 3D point cloud data can then be integrated with radiation source information to achieve 3D localization of the source. However, these existing radiation imaging technologies cannot calculate the intensity of the radiation source. They simply attach the hotspot image to the 3D scene as a texture, without further calculating the activity of the hotspot.

[0003] Given that current radiation imaging technology cannot provide the intensity value of the source term, the present invention proposes a quantitative imaging method for radioactive sources based on coded aperture technology and lidar, which can perform long-distance imaging positioning and activity calculation of radioactive sources, thereby achieving accurate measurement of radioactive sources. Summary of the Invention

[0004] As a supplement to existing radioactive source imaging methods and a methodological innovation, the present invention aims to address the problem that existing radioactive source imaging methods can only provide the location of the radiation source but not the intensity (activity or surface dose rate) of the radiation source. By proposing a quantitative imaging method based on coded aperture technology and lidar, it can calculate the activity of radioactive sources from a distance, providing a new methodological approach for imaging and intensity measurement of radioactive sources. The quantitative radioactive source imaging method proposed in this invention can simultaneously locate the radiation hotspot of the radiation source and provide its activity value. Compared with existing radioactive imaging methods, it also provides absolute intensity information of the radiation source, enabling the indication of the radiation source location and activity calculation at a distance.

[0005] The present invention proposes a radioactive source quantitative imaging method based on coded aperture technology. The flow chart of the method is as follows: Figure 1 As shown, the specific implementation steps include:

[0006] 1. The target area is imaged using a coded aperture gamma camera, and the coded position projection of the radiation source obtained by the position-sensitive imaging detector in the camera is decoded and calculated to obtain a coded reconstructed image.

[0007] 2. Perform filtering calculation on the encoded reconstructed image to shield the image noise, obtain the radiation hotspots in the image and generate a radiation hotspot image.

[0008] 3. Identify the full energy peak of the observed radioactive source based on the energy information (i.e., energy spectrum) obtained by the imaging detector in the gamma camera, and calculate the dose rate information at the position of the gamma camera based on the energy spectrum.

[0009] 4. Use the lidar to obtain three-dimensional point cloud information of the imaging area, and use the optical camera to obtain optical video information of the imaging area. Then, make azimuth correspondence between the pixels in the radiation hotspot image, optical video, and three-dimensional point cloud. The radiation hotspot pixels are first superimposed on the optical pixels. The fused pixels of the two are then superimposed on the pixels of the three-dimensional point cloud to achieve matching fusion of the three pixels. The final image is a three-dimensional scene with radiation hotspots.

[0010] 5. The distance information between the radiation source corresponding to the radiation hotspot and the gamma camera is obtained through the radar 3D point cloud pixels corresponding to the radiation hotspot pixels. Combined with the radiation source energy information and dose rate information obtained by the gamma camera, the activity value of the radiation source is inverted and calculated.

[0011] The technical solution of the present invention is:

[0012] A radioactive source quantitative imaging method based on coded aperture technology and laser radar, comprising the following steps:

[0013] 1) Imaging the target area with a coded aperture gamma camera, obtaining the coded projection and energy spectrum of the radioactive source in the target area, and decoding the coded projection to obtain a coded reconstructed image O;

[0014] 2) Obtaining radiation hot spots in the encoded reconstructed image O to generate a radiation hot spot image O_rad;

[0015] 3) identifying the full energy peak E of the radiation source according to the energy spectrum, and calculating the dose rate Dose_rate at the location of the coded aperture gamma camera according to the energy spectrum;

[0016] 4) Acquire three-dimensional point cloud information of the target area through a laser radar, and acquire optical video information of the target area through an optical camera; then, match and fuse the pixels in the radiation hotspot image O_rad, the optical video information, and the three-dimensional point cloud information according to the azimuth angle to obtain a three-dimensional scene image O_final with the radiation hotspot;

[0017] 5) searching for the three-dimensional point cloud pixels corresponding to the radiation hotspot in the three-dimensional scene graph O_final; obtaining the distance D between the radiation source and the coded aperture gamma camera based on the three-dimensional point cloud pixels corresponding to the radiation hotspot;

[0018] 6) According to The activity value Activity of the radioactive source corresponding to the radiation hotspot is calculated; wherein mfp is the mean free path, and B(E,mfp) represents the accumulation factor.

[0019] Furthermore, the encoded and reconstructed image O is filtered to obtain radiation hot spots in the encoded and reconstructed image O.

[0020] Furthermore, the method for obtaining the radiation hotspot in the encoded and reconstructed image O is as follows: perform histogram statistics on the pixel values ​​of the encoded and reconstructed image O to obtain the frequency distribution of the pixel values; obtain the maximum pixel value PEAK(O) and the minimum pixel value MIN(O) of the encoded and reconstructed image O; calculate the sum of the pixel values ​​SUM(O) of the encoded and reconstructed image O; then calculate the average value NOISE_MEAN(O) of the image noise, the cutoff threshold of the radiation hotspot is CUT(O), and judge whether each pixel value in the encoded and reconstructed image O is greater than CUT(O) in turn, and take the pixels greater than CUT(O) as the radiation hotspot.

[0021] Further, CUT(O)=|MIN(O)|+NOISE_MEAN(O); N is the total number of pixels of the encoded reconstructed image O.

[0022] Furthermore, the method for obtaining the three-dimensional scene graph O_final with the radiation hotspot is as follows: matching and aligning the fields of view of the radiation hotspot image O_rad, the optical video image and the three-dimensional point cloud image, and then making the radiation hotspot pixels of the radiation hotspot image O_rad transparent and fitting them on the corresponding pixels in the optical video image to obtain the image O_rad_opt after the radiation hotspot image O_rad and the optical video image are fused; selecting the field of view scanning angle range of the laser radar to cover the imaging field of view range of the radiation hotspot image O_rad, and the angular resolution of the three-dimensional point cloud pixels is greater than the angular resolution of the radiation hotspot image O_rad; and then superimposing the pixels of the image O_rad_opt on the corresponding pixels of the three-dimensional point cloud image to obtain the three-dimensional scene graph O_final.

[0023] Furthermore, the mean free path T(E) represents the linear energy conversion coefficient, and its value is determined by the energy E.

[0024] A radioactive source quantitative imaging system based on coded aperture technology and laser radar, characterized by comprising a coded aperture gamma camera, a laser radar, an optical camera and a data processing unit;

[0025] The coded aperture gamma camera is used to image the target area, obtain the coded projection and energy spectrum of the radioactive source in the target area, and send them to the data processing unit;

[0026] The laser radar obtains three-dimensional point cloud information of the target area and sends it to the data processing unit;

[0027] The optical camera acquires optical video information of the target area and sends the information to the data processing unit;

[0028] The data processing unit is used to decode and calculate the coded projection to obtain a coded reconstructed image O, then obtain the radiation hotspot in the coded reconstructed image O to generate a radiation hotspot image O_rad; and identify the full energy peak E of the radiation source according to the energy spectrum, and calculate the dose rate Dose_rate at the location of the coded aperture gamma camera according to the energy spectrum; and match and fuse the pixels in the radiation hotspot image O_rad, optical video information, and three-dimensional point cloud information according to the azimuth angle to obtain a three-dimensional stereoscopic scene graph O_final with the radiation hotspot; search for the three-dimensional point cloud pixels corresponding to the radiation hotspot in the three-dimensional stereoscopic scene graph O_final, and then obtain the distance D between the radiation source and the coded aperture gamma camera according to the three-dimensional point cloud pixels corresponding to the radiation hotspot; and then calculate the distance D between the radiation source and the coded aperture gamma camera according to the three-dimensional point cloud pixels corresponding to the radiation hotspot. The activity value Activity of the radioactive source corresponding to the radiation hotspot is calculated; wherein mfp is the mean free path, and B(E,mfp) represents the accumulation factor.

[0029] Furthermore, the method for the data processing unit to obtain the radiation hotspot in the encoded and reconstructed image O is: to perform histogram statistics on the pixel values ​​of the encoded and reconstructed image O to obtain the frequency distribution of the pixel values; to obtain the maximum pixel value PEAK(O) and the minimum pixel value MIN(O) of the encoded and reconstructed image O; to calculate the sum of the pixel values ​​SUM(O) of the encoded and reconstructed image O; and then to calculate the average value NOISE_MEAN(O) of the image noise, the cutoff threshold of the radiation hotspot is CUT(O), and to determine in turn whether each pixel value in the encoded and reconstructed image O is greater than CUT(O), and to take the pixels greater than CUT(O) as the radiation hotspot.

[0030] Furthermore, the data processing unit obtains the three-dimensional scene graph O_final with the radiation hotspot by: matching and aligning the fields of view of the radiation hotspot image O_rad, the optical video image and the three-dimensional point cloud image, and then making the radiation hotspot pixels of the radiation hotspot image O_rad transparent and superimposing them on the corresponding pixels in the optical video image to obtain the image O_rad_opt after the radiation hotspot image O_rad and the optical video image are fused; the field of view scanning angle range of the laser radar is selected to cover the imaging field of view range of the radiation hotspot image O_rad, and the angular resolution of the three-dimensional point cloud pixels is greater than the angular resolution of the radiation hotspot image O_rad; and then the pixels of the image O_rad_opt are superimposed on the corresponding pixels of the three-dimensional point cloud image to obtain the three-dimensional scene graph O_final.

[0031] Compared with the prior art, the present invention has the following positive effects:

[0032] 1) Existing radiation imaging technology can generally only provide the hotspot orientation and relative intensity value of the radiation source. The quantitative imaging method of radiation sources based on coded aperture technology and lidar proposed in this invention realizes accurate imaging measurement of radiation sources at long distances. The distance information of the radiation source is obtained by lidar to further calculate the activity value of the radiation source, thus realizing quantitative imaging of the radiation source.

[0033] 2) The proposed quantitative imaging method for radioactive sources based on coded aperture technology is similar to the existing imaging method in terms of workflow. It does not require multi-position or multi-angle measurements of the radioactive source to calculate its activity value. Instead, a single imaging operation at a fixed point is required to achieve imaging positioning and activity calculation of the radioactive source, without adding additional workload.

[0034] As a supplement to the radiation imaging method of radioactive sources, the present invention proposes a quantitative imaging method based on coded aperture technology and lidar that can perform long-distance activity calculation of radioactive sources. Compared with the original radiation imaging method, it provides more absolute intensity information of the radioactive source, and can realize the direction indication and activity calculation of the radioactive source at a long distance. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 Flow chart of the method of the present invention.

[0036] Figure 2 This is the energy spectrum obtained by sampling the imaging detector.

[0037] Figure 3 is the projection image of the imaging detector.

[0038] Figure 4 To reconstruct the image.

[0039] Figure 5 is the frequency distribution of pixel values.

[0040] Figure 6 It is the radiation hotspot image after interpolation and smoothing.

[0041] Figure 7 This is the fusion flow chart.

[0042] Figure 8 is the fused image. DETAILED DESCRIPTION

[0043] The present invention will be described in further detail below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0044] like Figure 1 As shown, the steps of the present invention include:

[0045] 1. Obtain the coded projection of the radiation source through a coded aperture gamma camera, and decode the coded projection to obtain a coded reconstructed image.

[0046] 1.1 Select a typical coded aperture gamma camera. Its key components include a code plate and an imaging detector. The code plate's code pattern can be MURA RANK-19, and the imaging detector is a pixel-type position-sensitive detector with a 19×19 pixel array. The imaging field of view is 40 degrees. Radiation emitted by a distant radiation source is modulated by the aperture of the code plate, forming a projection on the detector that corresponds to fluctuations in the corresponding counts. The mathematical form of the projection is a 19×19 two-dimensional array D.

[0047] 1.2 Perform convolution calculation on the projection array D and the decoding matrix G to obtain the encoded reconstructed image O. The mathematical form of the decoding matrix and the reconstructed image is also a 19×19 two-dimensional array. The convolution calculation formula is as follows:

[0048]

[0049] The “%” symbol indicates remainder calculation.

[0050] 2. Perform filtering calculation on the encoded reconstructed image O to shield the image noise and obtain the radiation hotspots in the image.

[0051] 2.1 For ease of explanation, assume the following imaging scenario: a Cs-137 radiation source with an activity of 1 mCi is placed 3 meters away from the gamma camera, the imaging detector is a 19×19 pixel array of CsI scintillator, and the energy spectrum obtained by the imaging detector sampling for 1 minute is as follows Figure 2 As shown, the imaging detector projection D is as follows Figure 3 As shown, the encoded reconstructed image O is as follows Figure 4 shown.

[0052] 2.2 The 361 pixel values ​​of the coded reconstructed image O are counted in histogram to obtain the frequency distribution of the pixel values, as shown in Figure 5 As shown, the maximum pixel value PEAK(O) and the minimum pixel value MIN(O) of the coded reconstructed image O are obtained, the sum of the pixel values ​​SUM(O) of the coded reconstructed image O is calculated, and then the average value of the image noise NOISE_MEAN(O) is calculated. The cutoff threshold of the radiation hotspot is CUT(O), and the 361 pixel values ​​in the coded reconstructed image O are judged in turn. Pixels larger than CUT(O) are regarded as radiation hotspots. Figure 6 The radiation hotspot image O_rad is obtained after deducting image noise and image interpolation and smoothing processing is performed.

[0053] The calculation formulas for NOISE_MEAN(O) and CUT(O) are:

[0054]

[0055] CUT(O)=|MIN(O)|+NOISE_MEAN(O)

[0056] 3. Find the full energy peak of the radiation source based on the energy spectrum obtained by the gamma camera, and calculate the dose rate information at the location of the gamma camera based on the energy spectrum.

[0057] 3.1 According to Figure 2 From the energy spectrum shown, we can obtain that the total energy peak position E is 662keV.

[0058] 3.2 According to Figure 2The energy spectrum shown can be used to calculate the average dose rate Dose_rate (0.31 μSv / h) obtained by the imaging detector within 1 minute. The dose rate calculation method can be implemented by referring to the common energy-dose rate conversion algorithm.

[0059] 4. Obtain the 3D point cloud information of the imaging area through the laser radar, obtain the optical video information of the imaging area through the optical camera, and perform pixel matching to fuse the three image information of the radiation hotspot image O_rad, optical video, and 3D point cloud to obtain the final fused image of the three messengers. The fusion principle of radiation imaging, optical video, and 3D point cloud is as follows: Figure 7 shown.

[0060] 4.1 Match the fields of view of the radiation hotspot image O_rad, the optical video image, and the radar point cloud image, that is, the field of view frames correspond to each other. Then, make the radiation hotspot pixels of the 19×19 radiation hotspot image O_rad transparent, and then fit and superimpose them on the corresponding pixels in the optical video image to obtain the image O_rad_opt that is the result of the fusion of the radiation hotspot image and the optical video image.

[0061] 4.2 The field of view scanning angle range of the selected lidar needs to cover the imaging field of view of the radiation image, and the angular resolution of the point cloud pixels must be better than that of the radiation image. The O_rad_opt image pixels are superimposed on the radar point cloud image pixels to obtain the final radiation, optical and radar three-messenger fusion image O_final, as shown in Figure 8 shown.

[0062] 5. Obtain the distance information corresponding to the radiation hotspot, combine it with the radiation source energy information and dose rate information obtained by the gamma camera, and inversely calculate the activity value of the radiation source.

[0063] 5.1 Find the point cloud pixel corresponding to the radiation hotspot pixel Pixel in the fused image O_final. Based on the distance information D (3m) corresponding to the point cloud pixel and the peak energy information E (662keV), the dose rate information calculated by the imaging detector in step 3.2 is Dose_rate (0.31μSv / h).

[0064] 5.2 Calculate the activity value (unit: mCi) of the radioactive source corresponding to the radiation hotspot using the following formula:

[0065]

[0066] Where mfp represents the mean free path, T(E) represents the linear energy conversion coefficient, whose value is determined by the energy E. B(E,mfp) represents the accumulation factor, whose value is determined by the energy E and the mean free path mfp. C(E) represents the energy-kerma conversion coefficient, whose value is determined by the energy E. The constant 0.76 is the fixed unit conversion coefficient. The above coefficients are calculated by lookup and interpolation. Table 1 shows the coefficient table for calculating the radiation hotspot activity.

[0067] Table 1 is the calculation coefficient table of radiation hotspot activity

[0068]

[0069] First, the mean free path mfp is calculated by interpolation. The ray energy E is 662keV, and the corresponding T(E) is:

[0070]

[0071]

[0072] Then the energy kerma conversion coefficient C(E) is calculated by interpolation:

[0073]

[0074] Then calculate the cumulative factor B(E,mfp) by interpolation:

[0075]

[0076] Therefore, the activity value of the radioactive source (unit: mCi) is calculated as:

[0077]

[0078] The above results show that the radiation hotspot quantitative imaging technology proposed in the present invention can image the radioactive source at a long distance and obtain its activity value.

[0079] In summary, the above is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A method for quantitative imaging of radioactive sources based on coded aperture technology and lidar, comprising the following steps: 1) Imaging the target area with a coded aperture gamma camera, obtaining the coded projection and energy spectrum of the radioactive source in the target area, and decoding the coded projection to obtain a coded reconstructed image O; 2) Obtain radiation hot spots in the coded reconstructed image O and generate a radiation hot spot image O_rad; the method for obtaining the radiation hot spots in the coded reconstructed image O is as follows: perform histogram statistics on the pixel values ​​of the coded reconstructed image O to obtain the frequency distribution of the pixel values; obtain the maximum pixel value PEAK(O) and the minimum pixel value MIN(O) of the coded reconstructed image O; calculate the sum of the pixel values ​​SUM(O) of the coded reconstructed image O; then calculate the average value NOISE_MEAN(O) of the image noise, the cutoff threshold of the radiation hot spot is CUT(O), and determine whether each pixel value in the coded reconstructed image O is greater than CUT(O), and use pixels greater than CUT(O) as the radiation hot spots; 3) identifying the full energy peak E of the radiation source according to the energy spectrum, and calculating the dose rate Dose_rate at the location of the coded aperture gamma camera according to the energy spectrum; 4) Acquire the three-dimensional point cloud information of the target area through the laser radar, and acquire the optical video information of the target area through the optical camera; then match and fuse the pixels in the radiation hotspot image O_rad, the optical video information, and the three-dimensional point cloud information according to the azimuth angle to obtain the three-dimensional stereo scene graph O_final with the radiation hotspot; the method for obtaining the three-dimensional stereo scene graph O_final with the radiation hotspot is as follows: match and align the fields of view of the radiation hotspot image O_rad, the optical video image, and the three-dimensional point cloud image, and then make the radiation hotspot pixels of the radiation hotspot image O_rad transparent and superimpose them on the corresponding pixels in the optical video image to obtain the image O_rad_opt after the radiation hotspot image O_rad and the optical video image are fused; select the field of view scanning angle range of the laser radar to cover the imaging field of view range of the radiation hotspot image O_rad, and the angular resolution of the three-dimensional point cloud pixels is greater than the angular resolution of the radiation hotspot image O_rad; then superimpose the pixels of the image O_rad_opt on the corresponding pixels of the three-dimensional point cloud image to obtain the three-dimensional stereo scene graph O_final; 5) searching for the three-dimensional point cloud pixels corresponding to the radiation hotspot in the three-dimensional scene graph O_final; obtaining the distance D between the radiation source and the coded aperture gamma camera based on the three-dimensional point cloud pixels corresponding to the radiation hotspot; 6) According to The activity value Activity of the radioactive source corresponding to the radiation hotspot is calculated; wherein mfp is the mean free path, B(E,mfp) represents the accumulation factor, and C(E) represents the energy-kerma conversion coefficient.

2. The method according to claim 1, characterized in that The encoded and reconstructed image O is filtered to obtain radiation hot spots in the encoded and reconstructed image O.

3. The method according to claim 2, characterized in that CUT(O)=|MIN(O)|+NOISE_MEAN(O); N is the total number of pixels of the encoded reconstructed image O.

4. The method according to claim 1 or 2, characterized in that The mean free path T(E) represents the linear energy conversion coefficient, and its value is determined by the energy E.

5. A radioactive source quantitative imaging system based on coded aperture technology and laser radar, characterized in that: Includes a coded aperture gamma camera, lidar, optical camera, and data processing unit; The coded aperture gamma camera is used to image the target area, obtain the coded projection and energy spectrum of the radioactive source in the target area, and send them to the data processing unit; The laser radar obtains three-dimensional point cloud information of the target area and sends it to the data processing unit; The optical camera acquires optical video information of the target area and sends the information to the data processing unit; The data processing unit is used to decode and calculate the coded projection to obtain a coded reconstructed image O, then obtain the radiation hotspot in the coded reconstructed image O to generate a radiation hotspot image O_rad; and identify the full energy peak E of the radiation source according to the energy spectrum, and calculate the dose rate Dose_rate at the location of the coded aperture gamma camera according to the energy spectrum; and match and fuse the pixels in the radiation hotspot image O_rad, optical video information, and three-dimensional point cloud information according to the azimuth angle to obtain a three-dimensional stereoscopic scene graph O_final with the radiation hotspot; search for the three-dimensional point cloud pixels corresponding to the radiation hotspot in the three-dimensional stereoscopic scene graph O_final, and then obtain the distance D between the radiation source and the coded aperture gamma camera according to the three-dimensional point cloud pixels corresponding to the radiation hotspot; and then calculate the distance D between the radiation source and the coded aperture gamma camera according to the three-dimensional point cloud pixels corresponding to the radiation hotspot. Calculate the activity value Activity of the radioactive source corresponding to the radiation hotspot; where mfp is the mean free path, B(E,mfp) represents the accumulation factor, and C(E) represents the energy-kerma conversion coefficient; The method for obtaining the radiation hotspot in the coded reconstructed image O is as follows: performing histogram statistics on the pixel values ​​of the coded reconstructed image O to obtain the frequency distribution of the pixel values; obtaining the maximum pixel value PEAK(O) and the minimum pixel value MIN(O) of the coded reconstructed image O; calculating the sum of the pixel values ​​SUM(O) of the coded reconstructed image O; then calculating the average value NOISE_MEAN(O) of the image noise, with the cutoff threshold of the radiation hotspot being CUT(O), and successively determining whether each pixel value in the coded reconstructed image O is greater than CUT(O), and taking the pixels greater than CUT(O) as the radiation hotspot; The method for obtaining the three-dimensional scene graph O_final with the radiation hotspot is as follows: matching and aligning the fields of view of the radiation hotspot image O_rad, the optical video image and the three-dimensional point cloud image, and then making the radiation hotspot pixels of the radiation hotspot image O_rad transparent and fitting them on the corresponding pixels in the optical video image to obtain the image O_rad_opt after the radiation hotspot image O_rad and the optical video image are fused; selecting the field of view scanning angle range of the laser radar to cover the imaging field of view range of the radiation hotspot image O_rad, and the angular resolution of the three-dimensional point cloud pixels is greater than the angular resolution of the radiation hotspot image O_rad; and then superimposing the pixels of the image O_rad_opt on the corresponding pixels of the three-dimensional point cloud image to obtain the three-dimensional scene graph O_final.

Images