Radiation image processing methods, learned models, radiation image processing modules, radiation image processing programs, and radiation image processing systems

Abstract

The present invention provides a radiographic image processing method, a learned model, a radiographic image processing module, a radiographic image processing program, and a radiographic image processing system for effectively removing noise from radiographic images. The control device (20) includes: an acquisition unit (201) that uses an image acquisition device (1) to irradiate an object (F) with X-rays and capture X-rays passing through the object (F) to acquire X-ray transmitted images of a fixture and the object (F); a determination unit (202) that determines the image characteristics of the X-ray transmitted image of the fixture; a selection unit (204) that selects a learned model (206) from a plurality of learned models (206) pre-constructed using image data through machine learning based on the image characteristics; and a processing unit (205) that uses the selected learned model (206) to perform image processing to remove noise from the X-ray transmitted image of the object (F).

Description

Technical Field

[0001] One aspect of the implementation relates to a radiation image processing method, a learned model, a radiation image processing module, a radiation image processing program, and a radiation image processing system. Background Technology

[0002] Currently, methods for noise removal using image data and learned models constructed through machine learning such as deep learning are known (see, for example, Patent Document 1 below). According to this method, since noise from image data is automatically removed, objects can be observed with high precision.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2019-91393 Summary of the Invention

[0006] The technical problem that the invention aims to solve

[0007] In existing methods as described above, when using radiation images generated by passing X-rays or other radiation through an object as the object, noise removal is sometimes insufficient. For example, there is a tendency that the relationship between image brightness and noise can easily vary depending on the conditions of the radiation source, such as the X-ray source, and the type of filter used, making it difficult to effectively remove noise.

[0008] Therefore, one aspect of the implementation is made in view of the above-mentioned problems, which is to provide a radiographic image processing method, a learned model, a radiographic image processing module, a radiographic image processing program, and a radiographic image processing system that can effectively remove noise from radiographic images.

[0009] Technical means to solve the problem

[0010] One aspect of the implementation method for radiographic image processing includes: acquiring a radiographic image of a fixture by using a system that irradiates an object with radiation and captures a radiographic image of the radiation passing through the object; determining image characteristics of the radiographic image of the fixture; selecting a learned model from a plurality of learned models pre-constructed using image data through machine learning based on the image characteristics; acquiring a radiographic image of the object using the system; and performing image processing to remove noise from the radiographic image of the object using the selected learned model.

[0011] Alternatively, another aspect of the implementation method may be a learned model used in the above-described radiation image processing method, which is constructed using image data through machine learning, enabling the processor to perform image processing to remove noise from the radiation image of the object.

[0012] Alternatively, the radiation image processing module in other aspects of the implementation includes: an acquisition unit that uses a system that irradiates an object with radiation and captures a radiation image by photographing the radiation passing through the object to acquire radiation images of a fixture and the object; a determination unit that determines the image characteristics of the radiation image of the fixture; a selection unit that selects a learned model from a plurality of learned models that have been pre-constructed using image data through machine learning based on the image characteristics; and a processing unit that uses the selected learned model to perform image processing to remove noise from the radiation image of the object.

[0013] Alternatively, in other aspects of the implementation, the radiation image processing program functions as an acquisition unit, a determination unit, a selection unit, and a processing unit. The acquisition unit uses a system that irradiates an object with radiation and captures a radiation image of the object by photographing the radiation passing through the object to acquire radiation images of a fixture and the object. The determination unit determines the image characteristics of the radiation image of the fixture. The selection unit selects a learned model from multiple learned models that have been pre-constructed using image data through machine learning based on the image characteristics. The processing unit uses the selected learned model to perform image processing to remove noise from the radiation image of the object.

[0014] Alternatively, other aspects of the implementation of the radiation image processing system include: the radiation image processing module described above; a source that irradiates a target with radiation; and an imaging device that captures the radiation passing through the target and obtains a radiation image.

[0015] Based on one or more of the aforementioned aspects, the image characteristics of the radiographic image of the fixture are determined. Based on these image characteristics, a pre-built, fully learned model for noise removal is selected. Thus, the characteristics of the radiographic image, which vary due to factors such as the conditions of the system's radiographic source, can be estimated. The fully learned model selected based on this estimation is then used for noise removal. Therefore, noise removal corresponding to the relationship between the brightness and noise of the radiographic image can be achieved. As a result, noise in the radiographic image can be effectively removed.

[0016] The effects of the invention

[0017] According to the implementation method, noise in the radiographic image of an object can be effectively removed. Attached Figure Description

[0018] Figure 1This is a schematic structural diagram of the image acquisition device 1 according to the embodiment.

[0019] Figure 2 It is shown Figure 1 A block diagram illustrating an example of the hardware structure of the control device 20.

[0020] Figure 3 It is shown Figure 1 A block diagram of the functional structure of the control device 20.

[0021] Figure 4 It is shown that it is used for Figure 3 The teaching data, i.e., image data, is an example of the learning data used to construct Model 206.

[0022] Figure 5 It is shown that it is used for Figure 3 The flowchart shows the sequence of creating the teaching data, i.e., the image data, for the completed model 206.

[0023] Figure 6 It is shown Figure 3 An example of an X-ray transmission image of the object analyzed by the determination unit 202.

[0024] Figure 7 It is shown Figure 3 An example of a thickness-brightness characteristic curve obtained by the determination unit 202.

[0025] Figure 8 It is shown Figure 3 An example of a brightness-SNR characteristic curve obtained by the determination unit 202.

[0026] Figure 9 It is shown that it is used by Figure 3 An example of an X-ray transmission image used to evaluate the resolution of the determination unit 202.

[0027] Figure 10 It is used to explain by Figure 3 The selection function of the learned model based on image characteristics is executed by the selection unit 204.

[0028] Figure 11 It is shown that it is used by Figure 3 A perspective view of an example of the structure of the fixture used for evaluating the brightness-to-noise ratio in the selection section 204.

[0029] Figure 12 It is shown that Figure 11 The image shows a noise-removed X-ray transmitted image obtained using a fixture as the object.

[0030] Figure 13This is a flowchart illustrating the sequence of observation processing using image acquisition device 1.

[0031] Figure 14 This is a diagram showing an example of X-ray transmitted images before and after noise removal processing acquired by image acquisition device 1.

[0032] Figure 15 This is a diagram showing an example of X-ray transmitted images before and after noise removal processing acquired by image acquisition device 1.

[0033] Figure 16 This is a top view showing the manner in which the fixture for the image acquisition device 1 is arranged.

[0034] Figure 17 This is a top view showing the manner in which the fixture for the image acquisition device 1 is arranged.

[0035] Figure 18 This is a top view showing the manner in which the fixture for the image acquisition device 1 is arranged. Detailed Implementation

[0036] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Furthermore, in the description, the same reference numerals are used for the same elements or elements having the same function, and repeated descriptions are omitted.

[0037] Figure 1 This is a structural diagram of the radiation image processing system, i.e., the image acquisition device 1, according to this embodiment. Figure 1 As shown, the image acquisition device 1 is an apparatus that irradiates an object F being transported along the transport direction TD with X-rays (radiation) and acquires an X-ray transmission image (radiation image) of the object F based on the X-rays passing through it. The image acquisition device 1 uses the X-ray transmission image to perform foreign object inspection, weight inspection, and inspection of the object F. Examples of its applications include food inspection, carry-on baggage inspection, substrate inspection, battery inspection, and material inspection. The image acquisition device 1 is configured to include: a belt conveyor (transportation mechanism) 60, an X-ray irradiator (radiation source) 50, an X-ray inspection camera (image capture device) 10, a control device (radiation image processing module) 20, a display device 30, and an input device 40 for various inputs. Furthermore, the radiation image in embodiments of the present invention is not limited to X-ray images, but also includes images generated by electromagnetic radiation other than X-rays such as gamma rays.

[0038] The belt conveyor 60 has a belt section carrying an object F. By moving this belt section along the conveying direction TD, the object F is conveyed at a predetermined conveying speed along the conveying direction TD. The conveying speed of the object F is, for example, 48 m / min. The belt conveyor 60 can change the conveying speed as needed, for example, to 24 m / min or 96 m / min. In addition, the belt conveyor 60 can appropriately change the height position of the belt section to change the distance between the X-ray irradiator 50 and the object F. Furthermore, the object F conveyed by the belt conveyor 60 can include, for example, various items such as edible meat, seafood, crops, snacks and other food products, rubber products such as tires, resin products, metal products, mineral resources and other materials, waste, and electronic components or electronic boards. The X-ray irradiator 50 is a device that irradiates (outputs) the object F with X-rays as an X-ray source. The X-ray irradiator 50 is a point source that diffuses X-rays within a predetermined angle range along a certain irradiation direction. An X-ray irradiator 50 is positioned above the belt conveyor 60, with the X-ray irradiation direction facing the belt conveyor 60 and the diffused X-rays reaching the entire width direction (the direction intersecting the transport direction TD) of the object F. Furthermore, the X-ray irradiator 50 defines a predetermined segment in the length direction of the object F (parallel to the transport direction TD) as the irradiation range, and the object F is transported by the belt conveyor 60 in the transport direction TD, thereby irradiating the entire length direction of the object F with X-rays. The X-ray irradiator 50 uses a control device 20 to set the tube voltage and tube current, irradiating the belt conveyor 60 with X-rays of a predetermined energy and radiation dose corresponding to the set tube voltage and tube current. Additionally, a filter 51 is provided near the belt conveyor 60 side of the X-ray irradiator 50 to allow the X-rays to pass through a predetermined wavelength band. The filter 51 is not necessarily required and may sometimes be omitted.

[0039] The X-ray inspection camera 10 detects X-rays that pass through the object F from the X-rays irradiated by the X-ray irradiator 50, and outputs a signal based on these X-rays. The X-ray inspection camera 10 is a dual-path X-ray camera with two sets of X-ray detection structures. In the image acquisition apparatus 1 of this embodiment, X-ray transmission images are generated based on the X-rays detected by each path (first path and second path) of the dual-path X-ray camera. Furthermore, by averaging or adding the two generated X-ray transmission images, a clear (bright) image can be obtained with less X-rays compared to the case where an X-ray transmission image is generated based on X-rays detected by a single path.

[0040] The X-ray inspection camera 10 includes: a filter 19; scintillators 11a and 11b; line scan cameras 12a and 12b; a sensor control unit 13; amplifiers 14a and 14b; AD converters 15a and 15b; correction circuits 16a and 16b; output interfaces 17a and 17b; and an amplifier control unit 18. The scintillators 11a, line scan cameras 12a, amplifiers 14a, AD converters 15a, correction circuits 16a, and output interfaces 17a are electrically connected, forming a first circuit structure. The scintillators 11b, line scan cameras 12b, amplifiers 14b, AD converters 15b, correction circuits 16b, and output interfaces 17b are also electrically connected, forming a second circuit structure. The line scan cameras 12a of the first circuit and 12b of the second circuit are arranged along the transport direction TD. Furthermore, in the following description, the structure common to both the first and second circuits will be represented by the structure of the first circuit.

[0041] The scintillator 11a is fixed to the line scanning camera 12a by adhesive or other means, converting X-rays passing through the object F into scintillation light. The scintillator 11a outputs the scintillation light to the line scanning camera 12a. The filter 19 allows the X-rays of a specified wavelength band to pass through towards the scintillator 11a. The filter 19 is not necessarily required and may sometimes be omitted.

[0042] The line scan camera 12a detects the flashing light from the scintillator 11a and converts it into electrical charge, which is output as a detection signal (electrical signal) to the amplifier 14a. The line scan camera 12a has multiple line sensors arranged in a direction intersecting the transport direction TD. The line sensors are, for example, CCD (Charge Coupled Device) image sensors or CMOS (Complementary Metal-Oxide Semiconductor) image sensors, and contain multiple photodiodes.

[0043] The sensor control unit 13 controls the line scan cameras 12a and 12b to repeatedly capture X-rays passing through the same area of ​​the object F, with a predetermined detection cycle. The predetermined detection cycle can be set based on factors such as the distance between the line scan cameras 12a and 12b, the speed of the belt conveyor 60, the distance between the X-ray irradiator 50 and the object F on the belt conveyor 60 (FOD (Focus Object Distance)), and the distance between the X-ray irradiator 50 and the line scan cameras 12a and 12b (FDD (Focus Detector Distance)). Alternatively, the predetermined cycle can be set individually based on the pixel width of the photodiodes in the direction orthogonal to the pixel arrangement direction of each of the line scan cameras 12a and 12b. In this case, the offset (delay time) of the detection cycle between the line scan cameras 12a and 12b can be determined based on the distance between the line scan cameras 12a and 12b, the speed of the belt conveyor 60, the distance (FOD) between the X-ray irradiator 50 and the object F on the belt conveyor 60, and the distance (FDD) between the X-ray irradiator 50 and the line scan cameras 12a and 12b, and individual cycles can be set for each. The amplifier 14a amplifies the detection signal at a predetermined set amplification rate and generates an amplified signal, which is then output to the AD converter 15a. The set amplification rate is the amplification rate set by the amplifier control unit 18. The amplifier control unit 18 sets the set amplification rate of the amplifiers 14a and 14b based on predetermined shooting conditions.

[0044] The AD converter 15a converts the amplified signal (voltage signal) output from amplifier 14a into a digital signal and outputs it to the correction circuit 16a. The correction circuit 16a performs prescribed corrections, such as signal amplification, on the digital signal and outputs the corrected digital signal to the output interface 17a. The output interface 17a outputs the digital signal to the outside of the X-ray inspection camera 10. Figure 1 In this system, the AD converter, correction circuit, or output interface may exist individually, but they can also be integrated into one.

[0045] The control device 20 is, for example, a computer such as a PC (Personal Computer). The control device 20 generates an X-ray transmission image based on the digital signals (amplified signals) output from the X-ray inspection camera 10 (more specifically, output interfaces 17a and 17b). The control device 20 generates one X-ray transmission image by averaging or adding the two digital signals output from output interfaces 17a and 17b. The generated X-ray transmission image is output to the display device 30 after noise removal processing described later, and is displayed by the display device 30. Furthermore, the control device 20 controls the X-ray irradiator 50, the amplifier control unit 18, and the sensor control unit 13. In this embodiment, the control device 20 is a device independently installed outside the X-ray inspection camera 10, but it can also be integrated inside the X-ray inspection camera 10.

[0046] Figure 2 The hardware structure of the control device 20 is shown. For example... Figure 2 As shown, the control device 20 physically includes a computer, such as a CPU (Central Processing Unit) 101 as a processor, RAM (Random Access Memory) 102 or ROM (Read Only Memory) 103 as a recording medium, a communication module 104, and an input / output module 106, all electrically connected. Furthermore, the control device 20 may also include a display, keyboard, mouse, touch panel display, etc., as input devices 40 and display devices 30, and may also include data recording devices such as hard disk drives and semiconductor memory. Additionally, the control device 20 may be composed of multiple computers.

[0047] Figure 3 This is a block diagram showing the functional structure of the control device 20. The control device 20 includes: an acquisition unit 201, a determination unit 202, a selection unit 204, and a processing unit 205. Figure 3 The functional units of the control device 20 shown are implemented by loading a program (the radiographic image processing program of this embodiment) onto hardware such as the CPU 101 and RAM 102. Under the control of the CPU 101, the communication module 104 and the input / output module 106 are operated, and data is read from and written to the RAM 102. The CPU 101 of the control device 20 executes this computer program to make the control device 20 function as... Figure 3Each functional unit performs its function, sequentially executing the processing corresponding to the radiation image processing method described later. Furthermore, the CPU can be a standalone hardware unit, or it can be installed within a programmable logic device such as an FPGA, like a software processor. Similarly, RAM or ROM can be a standalone hardware unit, or it can be built into a programmable logic device such as an FPGA. All data required for executing the computer program, and all data generated by executing the computer program, are stored in built-in memory such as ROM 103, RAM 102, or memory media such as a hard disk drive.

[0048] Furthermore, the control device 20 pre-stores multiple learned models 206, which are read by the CPU 101, enabling the CPU 101 to perform noise removal processing on the X-ray transmitted image. The multiple learned models 206 are learning models pre-constructed using image data as teaching data and generated by machine learning. Machine learning methods include teaching-based learning, deep learning, reinforcement learning, and neural network learning. In this embodiment, as an example of a deep learning algorithm, a two-dimensional convolutional neural network described in the paper "Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising" by Kai Zhang et al. is used. The multiple learned models 206 can also be generated by an external computer and downloaded to the control device 20, or they can be generated within the control device 20.

[0049] exist Figure 4 The image data shown is an example of the teaching data used to construct the model 206 after learning. As teaching data, X-ray transmission images of patterns of various thicknesses, materials, and resolutions can be used. Figure 4The example shown is an X-ray transmission image of chicken. This image data can also be X-ray transmission images actually generated using image acquisition device 1 with various objects, or image data generated through simulation calculations. Regarding the X-ray transmission image, it can also be an image obtained using a device different from image acquisition device 1. Furthermore, X-ray transmission images can be combined with image data generated through simulation calculations. Multiple learned models 206 are pre-built using image data obtained with transmitted X-rays of different average energies and with known noise distributions. The average energy of the X-rays in the image data is preset to different values ​​by setting the operating conditions of the X-ray irradiator (radiation source) 50 of image acquisition device 1, or the imaging conditions of image acquisition device 1, or by setting the operating conditions or imaging conditions of the X-ray irradiator 50 during simulation calculations. That is, multiple learned models 206 are constructed using training images as training data through machine learning (construction step). The training images are X-ray images corresponding to the average energy of X-rays transmitted through the object F, calculated based on conditional information such as the operating conditions of the X-ray irradiator (radiation source) 50 representing the X-ray transmission image of the object F, or the imaging conditions of the X-ray detection camera 10. For example, in this embodiment, the multiple learned models 206 are constructed using various image data with average energies set to 10keV, 20keV, 30keV, ... and 10keV scale values ​​for multiple frames (e.g., 20,000 frames).

[0050] Figure 5 A flowchart showing the sequence of creating the teaching data, i.e., the image data, used to construct the model 206 after learning is completed.

[0051] The teaching data, i.e., image data (also called teaching image data), is created by the computer in the following order. First, an image of a structure with a specified structure (structure image) is created (step S101). For example, an image of a structure with a specified structure can also be created through simulation calculation. Alternatively, an X-ray image of a structure such as a chart with a specified structure can be obtained to create a structure image. Next, for a pixel selected from the multiple pixels constituting the structure image, the standard deviation of the pixel value, i.e., σ (sigma), is calculated (step S102). Then, a normal distribution (Poisson distribution) representing the noise distribution is set based on the σ value obtained in step S102 (step S103). In this way, teaching data for various noise conditions can be generated by setting a normal distribution based on the σ value. Next, a randomly set noise value is calculated according to the normal distribution set based on the σ value in step S103 (step S104). Furthermore, by appending the noise value obtained in step S104 to the pixel value of a pixel, pixel values ​​constituting the teaching data, i.e., image data, are generated (step S105). Steps S102 to S105 are then processed on multiple pixels constituting the structure image (step S106) to generate teaching image data that becomes teaching data (step S107). Additionally, if teaching image data is still needed, steps S101 to S107 are processed on other structure images (step S108) to generate other teaching image data that becomes teaching data. Furthermore, the other structure images can be images of structures with the same structure or images of structures with different structures.

[0052] Furthermore, multiple teaching data, i.e., image data, are needed for constructing the learned model 206. Additionally, images with low noise are preferred among the structural images; ideally, noise-free images are best. Therefore, if structural images are generated through simulation calculations, multiple noise-free images can be produced, making simulation calculations an effective method for generating structural images.

[0053] Below, return Figure 3 The functions of each functional unit of the control device 20 will be explained in detail.

[0054] The acquisition unit 201 acquires X-ray transmission images of a fixture and an object F irradiated with X-rays using the image acquisition device 1. The fixture is a flat plate with known thickness and material, and a known relationship between the average X-ray energy and X-ray transmittance, or a fixture with charts taken at various resolutions. That is, the acquisition unit 201 acquires the X-ray transmission image of the fixture taken by the image acquisition device 1 before observing the object F. Furthermore, the acquisition unit 201 acquires the X-ray transmission image of the object F taken by the image acquisition device 1 at a time after the model 206 has been learned, based on the X-ray transmission image of the fixture. However, the acquisition time of the X-ray transmission images of the fixture and the object F is not limited to the above; they can be acquired simultaneously or at opposite times.

[0055] The determination unit 202 determines the image characteristics of the X-ray transmission image of the fixture acquired by the acquisition unit 201. Specifically, the selection unit 204 determines the image characteristics of the X-ray transmission image, such as energy characteristics, noise characteristics, resolution characteristics, or frequency characteristics.

[0056] For example, when using a flat plate member of known thickness and material as a fixture, the determination unit 202 compares the brightness of the X-ray image transmitted through the fixture with the brightness of the X-ray image transmitted through air, and calculates the X-ray transmittance of one point (or an average of multiple points) of the fixture. For example, if the brightness of the X-ray image transmitted through the fixture is 5550 and the brightness of the X-ray image transmitted through air is 15000, the transmittance is calculated to be 37%. Furthermore, the determination unit 202 determines the average energy of the transmitted X-rays (e.g., 50 keV) estimated based on the transmittance of 37% as the energy characteristic of the X-ray transmitted image of the fixture.

[0057] In addition, the determination unit 202 can also analyze the characteristics of multiple points of the fixture with changes in thickness or material, as well as the energy characteristics of the X-ray transmission image of the fixture. Figure 6 This is a diagram showing an example of an X-ray transmission image of the object being analyzed by the determination unit 202. Figure 6 This is an X-ray transmission image of a fixture with a shape that varies in thickness in a stepped manner. The determination unit 202 selects multiple measurement areas (ROIs) with different thicknesses based on this X-ray transmission image, analyzes the average brightness of each of the multiple measurement areas, and obtains a thickness-brightness characteristic curve as an energy characteristic. Figure 7 The image shows an example of a thickness-brightness characteristic curve obtained by the determination unit 202.

[0058] Furthermore, the determination unit 202 can analyze the luminance value and noise of each of the multiple measurement areas as the noise characteristics of the X-ray transmission image of the fixture, and obtain a luminance-noise ratio characteristic curve as the noise characteristic. That is, the determination unit 202 selects multiple measurement areas (ROIs) with different thicknesses or materials based on the X-ray transmission image, analyzes the standard deviation and average luminance value of the multiple measurement areas (ROIs), and obtains a luminance-SNR (SN ratio) characteristic curve as the noise characteristic. At this time, the determination unit 202 calculates the SNR of each measurement area (ROI) using SNR = (average luminance value) ÷ (standard deviation of luminance value). Figure 8 The diagram shows an example of a luminance-SNR characteristic curve obtained by the determination unit 202. Alternatively, the determination unit 202 may obtain a noise characteristic curve with the vertical axis set to the standard deviation of the luminance value, instead of the aforementioned luminance-SNR characteristic curve, as the noise characteristic.

[0059] Furthermore, when using a fixture with a graph, the determination unit 202 can also obtain the resolution distribution of the X-ray transmission image of the fixture as a resolution characteristic. Moreover, the determination unit 202 has the function of obtaining the resolution characteristics for the image after applying multiple learned models 206 and performing noise removal processing on the X-ray transmission image of the fixture. Figure 9 The image shown is an example of an X-ray transmission image used for resolution evaluation. In this X-ray transmission image, a graph showing a stepwise variation of resolution along one direction is used as the subject of the photograph. The resolution of an X-ray transmission image can be determined using MTF (Modulation Transfer Function) or CTF (Contrast Transfer Function).

[0060] Refer again Figure 3 The selection unit 204 selects, from among the multiple learned models 206 stored in the control device 20, the learned model 206 ultimately used for noise removal processing of the X-ray transmission image of the object F, based on the image characteristics obtained by the determination unit 202. That is, the selection unit 204 compares the image characteristics determined by the determination unit 202 with the image characteristics determined based on the image data used to construct the multiple learned models 206, and selects the learned model 206 that is similar in both.

[0061] For example, the selection unit 204 selects a learned model 206, which is constructed from image data whose average energy is closest to the value of the average energy of transmitted X-rays determined by the determination unit 202.

[0062] Furthermore, similar to the determination method performed by the determination unit 202, the selection unit 204 obtains a thickness-brightness characteristic curve based on the image data used to construct multiple learned models 206, and selects the learned model 206 constructed from image data having characteristics closest to the thickness-brightness characteristic curve obtained for the jig as the final learned model 206. However, the image characteristics of the image data used to construct the learned model 206 can also be referenced to image characteristics calculated in advance outside the control device 20. In this way, by using the image characteristics obtained by setting multiple measurement areas, the most suitable learned model for noise removal of the X-ray transmission image of the object F can be selected. In particular, the differences in the X-ray spectrum or the effects of the filter during the measurement of the X-ray transmission image can be estimated with high precision.

[0063] Alternatively, the selection unit 204 may select the learned model 206 constructed from image data having a luminance-noise ratio characteristic that is closest to the luminance-noise ratio characteristic obtained by the determination unit 202 as the final learned model 206. However, the image characteristics of the image data used to construct the learned model 206 may also be obtained by the selection unit 204 based on the image data, or may refer to image characteristics calculated in advance outside the control device 20. Here, the selection unit 204 may also use the luminance-noise characteristic instead of the luminance-noise ratio characteristic as the noise characteristic to select the learned model 206. By using such a luminance-noise characteristic, the dominant noise factor (shot noise, readout noise, etc.) can be determined for each signal quantity detected by the X-ray detection camera 10 based on the slope of the curve of the region of each signal quantity, and the learned model 206 can be selected based on the determination result.

[0064] Figure 10 This diagram illustrates the selection function of the learned model based on image characteristics, performed by the selection unit 204. Figure 10 In section (a), the brightness-SNR characteristic curves G1, G2, and G3 of the various image data used for constructing multiple learned models 206 are shown. In section (b), in addition to these characteristic curves G1, G2, and G3, the brightness-SNR characteristic curve G of the X-ray transmitted image of the fixture is also shown. T In the characteristic curves G1, G2, G3, G... T In the case of an object, selection section 204 selects the one closest to the characteristic curve G. T The learned model 206 is constructed using the image data of the characteristic curve G2, which represents the characteristics of the model, and functions accordingly. During selection, the selection unit 204 selects the characteristic curves G1, G2, G3, and characteristic curve G...T In between, the SNR error of the brightness values ​​at regular intervals is calculated, and the root mean square error (RMSE) of these errors is calculated. The learned model 206 corresponding to the characteristic curves G1, G2, and G3 with the smallest RMSE is selected. Alternatively, when the selection unit 204 uses energy characteristics for selection, the learned model 206 can also be selected in the same way.

[0065] The selection unit 204 may also take the X-ray transmitted image of the fixture as the object, and select the learned model 206 for generating an image with relatively superior characteristics based on the characteristics of the image after noise removal processing is performed by applying multiple learned models.

[0066] For example, the selection unit 204 uses X-ray transmission images of a fixture that captures charts with various resolutions, applies multiple learned models 206 to these images, and evaluates the resolution characteristics of the resulting noise-removed image. Furthermore, the selection unit 204 selects the learned model 206 from the images whose resolution changes least before and after the noise removal process.

[0067] In addition to evaluating the resolution change mentioned above, the selection unit 204 can also evaluate the brightness-to-noise ratio characteristics of the noise-removed image and select the learned model 206 that generates the image with the highest brightness-to-noise ratio. Figure 11 The image shows an example of the structure of a fixture used for evaluating the luminance-noise ratio. For example, a fixture could be used in which foreign objects P2 of various materials and sizes are scattered in a component P1 whose thickness varies in a stepped manner along one direction. Figure 12 Showing Figure 11 The fixture obtains a noise-removed X-ray transmitted image of the object. The selection unit 204 selects an image region R1 containing the image of the foreign object P2 and an image region R2 containing the image of the foreign object P2 in the vicinity of region R1, and calculates the minimum brightness L of image region R1. MIN The average brightness L of image region R2 AVE The standard deviation L of the brightness of image region R2. SD Furthermore, the selection unit 204 uses the following formula:

[0068] CNR=(L AVE -L MIN ) / L SD

[0069] The brightness-to-noise ratio (CNR) is calculated. Then, the selection unit 204 calculates the brightness-to-noise ratio (CNR) for each of the X-ray transmission images after applying multiple learned models 206, and selects the learned model 206 that generates the X-ray transmission image with the highest brightness-to-noise ratio (CNR).

[0070] Alternatively, selection unit 204 may also base its selection on the average brightness L of image region R1. AVE_R1 The average brightness L of image region R2 AVE_R2 The standard deviation L of the brightness of image region R2. SD The following formula is used for calculation.

[0071] CNR=(L AVE_R1 -L MIN_R2 ) / L SD

[0072] The processing unit 205 applies the learned model 206 selected by the selection unit 204 to the X-ray transmission image obtained with the object F as the target, and performs noise removal image processing to generate an output image. The processing unit 205 then outputs the generated output image to a display device 30 or the like.

[0073] Next, the sequence of observation and processing of the X-ray transmitted image of the object F using the image acquisition apparatus 1 of this embodiment, i.e., the flow of the radiation image processing method of this embodiment, will be described. Figure 13 This is a flowchart illustrating the sequence of observation processes performed by the image acquisition device 1.

[0074] First, the operator (user) of the image acquisition device 1 sets the imaging conditions of the image acquisition device 1, such as the tube voltage of the X-ray irradiator 50 or the gain of the X-ray detection camera 10 (step S1). Next, a fixture is set in the image acquisition device 1, and the X-ray transmission image is acquired with the fixture as the object by the control device 20 (step S2). At this time, X-ray transmission images of various fixtures can also be acquired sequentially.

[0075] Correspondingly, the control device 20 determines the image characteristics (energy characteristics, noise characteristics, and resolution characteristics) of the X-ray transmission image of the fixture (step S3). Furthermore, the control device 20 applies multiple learned models 206 to the X-ray transmission image of the fixture and determines the image characteristics (resolution characteristics or brightness-noise ratio, etc.) of each X-ray transmission image after applying the multiple learned models 206 (step S4).

[0076] Next, the control device 20 selects the learned model 206 based on a comparison of the energy characteristics of the X-ray transmitted image of the fixture with the energy characteristics of the image data used to construct the learned model 206, and the degree of change in the resolution characteristics of the X-ray transmitted image of the fixture before and after the application of the learned model (step S5). Alternatively, the learned model 206 can also be selected based on a comparison of the noise characteristics of the X-ray transmitted image of the fixture with the noise characteristics of the image data used to construct the learned model 206, and the degree of change in the resolution characteristics of the X-ray transmitted image of the fixture before and after the application of the learned model. Alternatively, in step S5, instead of the above processing, the learned model 206 with the highest brightness-to-noise ratio (CNR) after applying the learned model can be selected.

[0077] Next, by setting the object F in the image acquisition device 1 and taking a picture of the object F, an X-ray transmission image of the object F is obtained (step S7). Then, by applying the finally selected learned model 206 to the X-ray transmission image of the object F by the control device 20, noise removal processing is performed on the X-ray transmission image as the object (step S8). Finally, the X-ray transmission image with noise removal processing, i.e., the output image, is output to the display device 30 by the control device 20 (step S9).

[0078] Based on the image acquisition apparatus 1 described above, the image characteristics of the X-ray image of the fixture are determined, and based on these image characteristics, a learned model for noise removal is selected from a pre-built learned model. Thus, the characteristics of the X-ray transmitted image, which vary due to the operating conditions of the X-ray irradiator 50 of the image acquisition apparatus 1, can be estimated. The learned model 206 selected based on this estimation result is then used for noise removal. Therefore, noise removal corresponding to the relationship between the brightness and noise of the X-ray transmitted image can be achieved. As a result, noise in the X-ray transmitted image can be effectively removed.

[0079] Generally, X-ray transmission images contain noise generated by X-rays. Increasing the X-ray dose to improve the signal-to-noise ratio (SN ratio) of the X-ray transmission image is also considered. However, in this case, increasing the X-ray dose increases the irradiation of the sensor, shortening its lifespan, and also shortens the lifespan of the X-ray source, making it difficult to simultaneously improve the SN ratio and extend its lifespan. In this embodiment, since it is not necessary to increase the X-ray dose, both improving the SN ratio and extending the lifespan can be achieved.

[0080] In this embodiment, during the selection of the learned model, the image characteristics of the X-ray transmission image of the fixture are compared with the image characteristics of the image data used to construct the learned model. Therefore, the learned model 206 constructed using image data corresponding to the image characteristics of the X-ray transmission image of the fixture is selected, thus effectively removing noise from the X-ray transmission image of the object F.

[0081] Furthermore, in this embodiment, the learned model is selected by applying image characteristics of multiple learned models 206 to the X-ray transmission image of the fixture. In this case, by actually applying the image characteristics of the X-ray transmission image of the fixture with multiple learned models 206, the learned model 206 is selected, thus effectively removing noise from the X-ray transmission image of the object F.

[0082] In particular, in this embodiment, energy characteristics or noise characteristics are used as image characteristics. In this case, a learned model 206 is selected by constructing an image whose characteristics are similar to the energy characteristics or noise characteristics of the X-ray transmission image of the fixture, which varies due to the imaging conditions of the image acquisition device 1. As a result, noise removal of the X-ray transmission image of the object F corresponding to the changes in the conditions of the image acquisition device 1 becomes possible.

[0083] In this embodiment, resolution characteristics or brightness-to-noise ratio are also used as image characteristics. Based on this structure, by applying the selected learned model 206, an X-ray transmission image with good resolution characteristics or brightness-to-noise ratio can be obtained. As a result, noise removal of the X-ray transmission image of the object becomes possible, corresponding to changes in the conditions of the image acquisition device 1.

[0084] exist Figure 14 and Figure 15 The image shows an example of X-ray transmission images before and after noise removal processing, acquired by the image acquisition device 1. Figure 14 and Figure 15 Images of cheese with foreign objects such as metal and glass are shown, as well as images of chicken with bones of various sizes remaining. The left side shows the image before noise removal, and the right side shows the image after noise removal. Therefore, it can be seen that this embodiment effectively removes noise from various objects.

[0085] The various embodiments of the present invention have been described above, but the present invention is not limited to the above embodiments. Modifications can be made without changing the spirit of the claims, or it can be applied to other embodiments.

[0086] For example, the X-ray inspection camera 10 has been described as a dual-line X-ray camera, but it is not limited to this. It can also be a single-line X-ray camera, a dual-energy X-ray camera, a TDI (Time Delay Integration) scanning X-ray camera, a multi-line X-ray camera with two or more lines, a two-dimensional X-ray camera, an X-ray flat panel sensor, an X-ray II, a direct conversion X-ray camera without a scintillator (a-Se, Si, CdTe, CdZnTe, TlBr, PbI2, etc.), or a camera that uses an optical lens that couples a scintillator lens. Furthermore, the X-ray inspection camera 10 can also be a camera tube sensitive to radiation, or a point sensor sensitive to radiation.

[0087] Furthermore, the image acquisition device 1 is not limited to the above-described embodiment; it can also be a radiographic image processing system, such as a CT (Computed Tomography) device, that takes images while keeping the object F stationary. Moreover, it can also be a radiographic image processing system that takes images while rotating the object F.

[0088] Furthermore, various types of fixtures can be used in the image acquisition apparatus 1 of the above embodiment. For example, a fixture such as... Figure 16 As shown, this is a jig with flat plate-shaped components P11, P12, P13, and P14 made of different materials arranged in a two-dimensional configuration. Alternatively, it can be as follows... Figure 17 As shown, the component has a one-dimensional, stepped shape and is arranged with flat components P21, P22, and P23 of different materials. Alternatively, part of the fixture may be open or notched, allowing the object F or something similar to object F to be photographed while photographing the fixture. The object F can be photographed while photographing the fixture, and the transmitted image of the object can be compared with the transmitted image of the fixture to select and learn the model. Furthermore, as... Figure 18 As shown in sections (a) to (c), a fixture with a diagram having boundary lines arranged in a manner that is parallel, oblique, or perpendicular to the conveying direction TD of the belt conveyor 60 can also be used.

[0089] In the above embodiments, the selection step is preferably performed by comparing image characteristics with image characteristics determined based on image data to select the learned model. In the above embodiments, it is preferable that the selection unit selects the learned model by comparing image characteristics with image characteristics determined based on image data. Therefore, by selecting the learned model constructed from image data corresponding to the image characteristics of the fixture's radiographic image, noise in the radiographic image of the object can be effectively removed.

[0090] Furthermore, in the determination step, it is preferable to determine the image characteristics of multiple images obtained as a result of applying multiple learned models to the radiographic image of the jig, and in the selection step, to select a learned model based on the image characteristics of the multiple images. Alternatively, it is preferable that the determination unit determines the image characteristics of the multiple images obtained as a result of applying multiple learned models to the radiographic image of the jig, and the selection unit selects a learned model based on the image characteristics of the multiple images. In this case, by selecting a learned model based on the image characteristics of the radiographic image of the jig to which multiple learned models have been applied, noise in the radiographic image of the object can be effectively removed.

[0091] Furthermore, preferably, the image characteristic is at least one of energy characteristics, noise characteristics, and frequency characteristics, and in the selection step, a learned model constructed from image data with similar image characteristics is selected. Furthermore, preferably, the image characteristic is at least one of energy characteristics, noise characteristics, and frequency characteristics, and the selection unit selects a learned model constructed from image data with similar image characteristics. In this case, a learned model constructed from an image whose characteristics are similar to at least one of the energy characteristics, noise characteristics, and frequency characteristics of the radiation image of the fixture, which varies due to the system, is selected. As a result, noise removal of the radiation image of the object corresponding to changes in system conditions becomes possible.

[0092] Furthermore, preferably, the image characteristic is resolution characteristic or brightness-to-noise ratio, and the selection step further includes: selecting a learned model for generating an image with relatively excellent resolution characteristics or brightness-to-noise ratio. Additionally, preferably, the image characteristic is resolution characteristic or brightness-to-noise ratio, and the selection unit selects a learned model for generating an image with relatively excellent resolution characteristics or brightness-to-noise ratio. According to this structure, by applying the selected learned model, a radiographic image with good resolution characteristics or brightness-to-noise ratio can be obtained. As a result, noise removal of the radiographic image of the object corresponding to changes in system conditions becomes possible.

[0093] [Industry availability]

[0094] The implementation method, which uses a radiographic image processing method, a learned model, a radiographic image processing module, a radiographic image processing program, and a radiographic image processing system for application, can effectively remove noise from radiographic images.

[0095] Symbol Explanation

[0096] 10…X-ray inspection camera (imaging device), 20…control device (radiation image processing module), 201…acquisition unit, 202…determination unit, 204…selection unit, 205…processing unit, 206…learned model, F…object, TD…transfer direction.

Claims

1. A method for processing radiation images, comprising: The step of obtaining a radiation image of a jig using a system that irradiates an object with radiation and captures a radiation image through the object with the radiation; The step of determining the image characteristics of the radiographic image of the fixture; Based on the image characteristics, the step of selecting a learned model from multiple learned models that have been pre-built using image data through machine learning. The steps of using the system to obtain a radiographic image of the object; and Using the selected learned model, perform image processing steps to remove noise from the radiation image of the object. The fixture is a flat plate-shaped component of known thickness and material, and for which the relationship between average X-ray energy and X-ray transmittance is known, or an object having charts taken at various resolutions. The multiple learned models are pre-built using image data obtained from X-rays with different average energies and known noise distribution as teaching data, and noise removal processing is performed to remove noise from the image data.

2. The radiation image processing method according to claim 1, wherein, In the selection step, the learned model is selected by comparing the image characteristics with the image characteristics determined based on the image data.

3. The radiation image processing method according to claim 1, wherein, In the determination step, image characteristics of multiple images obtained by applying the multiple learned models to the radiation image of the fixture are determined. In the selection step, the learned model is selected based on the image characteristics of the plurality of images.

4. The radiation image processing method according to claim 2, wherein, The image characteristic is at least one of energy characteristics, noise characteristics, and frequency characteristics. In the selection step, the learned model is selected from image data with similar image characteristics.

5. The radiation image processing method according to claim 3, wherein, The image characteristics mentioned are resolution characteristics or brightness-to-noise ratio. In the selection step, the learned model is selected to generate images with relatively good resolution characteristics or brightness-noise ratio.

6. The radiation image processing method according to any one of claims 1 to 5, wherein, The machine learning mentioned is deep learning.

7. A learned model, wherein, It is a learned model used in the radiation image processing method according to any one of claims 1 to 6. The learned model is constructed using image data through machine learning, enabling the processor to perform image processing to remove noise from the radiometric image of the object.

8. A radiation image processing module, comprising: The acquisition unit uses a system that irradiates an object with radiation and captures a radiation image by taking a picture of the radiation passing through the object, thereby acquiring radiation images of the fixture and the object; A determining unit that determines the image characteristics of the radiographic image of the fixture; The selection unit, based on the image characteristics, selects a learned model from multiple learned models pre-constructed using image data through machine learning; and The processing unit, using the selected learned model, performs image processing to remove noise from the radiation image of the object. The fixture is a flat plate-shaped component of known thickness and material, and the relationship between the average energy of X-rays and the X-ray transmittance is known, or an object having charts taken at various resolutions. The multiple learned models are pre-built using image data obtained from X-rays with different average energies and known noise distribution as teaching data, and noise removal processing is performed to remove noise from the image data.

9. The radiation image processing module according to claim 8, wherein, The selection unit selects the learned model by comparing the image characteristics with the image characteristics determined based on the image data.

10. The radiation image processing module according to claim 8, wherein, The determining unit determines the image characteristics of multiple images obtained by applying the multiple learned models to the radiation image of the fixture. The selection unit selects the learned model based on the image characteristics of the multiple images.

11. The radiation image processing module according to claim 9, wherein, The image characteristic is at least one of energy characteristics, noise characteristics, and frequency characteristics. The selection unit selects the learned model constructed from image data with similar image characteristics.

12. The radiation image processing module according to claim 10, wherein, The image characteristics mentioned are resolution characteristics or brightness-to-noise ratio. The selection unit selects the learned model for generating images with relatively excellent resolution characteristics or brightness-to-noise ratio.

13. The radiation image processing module according to any one of claims 8 to 12, wherein, The machine learning mentioned is deep learning.

14. A radiographic image processing program that enables a processor to function as an acquisition unit, a determination unit, a selection unit, and a processing unit. The acquisition unit uses a system that irradiates a target object with radiation and captures a radiation image by photographing the radiation transmitted through the target object, thereby acquiring radiation images of the fixture and the target object. The determining unit determines the image characteristics of the radiation image of the fixture. The selection unit, based on the image characteristics, selects a learned model from multiple learned models pre-constructed using image data through machine learning. The processing unit uses the selected learned model to perform image processing to remove noise from the radiation image of the object. The fixture is a flat plate-shaped component of known thickness and material, and the relationship between the average energy of X-rays and the X-ray transmittance is known, or an object having charts taken at various resolutions. The multiple learned models are pre-built using image data obtained from X-rays with different average energies and known noise distribution as teaching data, and noise removal processing is performed to remove noise from the image data.

15. A radiation image processing system, comprising: The radiation image processing module according to any one of claims 8 to 13; A source that irradiates the object with radiation; and An imaging device that captures radiation passing through the object to obtain an image of the radiation.

Images