CT examination equipment

Abstract

To provide a CT inspection device capable of generating more accurate CT sectional images with a smaller number of projection images. [Solution] The present invention relates to a CT inspection device, and is characterized by including a CT imaging unit that captures N projection images of an object to be inspected, and a CT image generating unit having an artificial intelligence-based AI model that receives the N projection images captured by the CT imaging unit and outputs a CT tomographic image of the object to be inspected. This makes it possible to generate a CT tomographic image with high image quality and resolution by capturing only a small number of projection images, for example, four projection images.

Description

[Technical Field]

[0001] The present invention relates to a CT examination apparatus, and more particularly to a CT examination apparatus that can generate a more accurate CT image with a smaller number of projection images. [Background technology]

[0002] Generally, small electronic components such as BGA (ball grid array) or CSP (chip scale package) are mounted by soldering on printed circuit boards (PCBs), which are incorporated as main components in electrical and electronic products such as home appliances and computers.

[0003] Therefore, when such a printed circuit board is assembled into an electric / electronic product set, the soldering condition of the mounted electronic components is inspected to determine whether it is good or bad.To inspect the soldering condition of such a printed circuit board, an X-ray inspection device is currently being used.

[0004] Such an X-ray inspection device for printed circuit boards is installed inside a cabinet manufactured to form an interior space for shielding, and irradiates X-rays onto a printed circuit board set at an inspection position, captures the projected image with a detector, and determines the quality of the soldering condition of the soldered parts of the printed circuit board based on the output image information.

[0005] Conventional X-ray inspection devices use a laminography-based tomographic imaging algorithm, which obtains cross-sectional layers for one image by synchronously rotating an X-ray generator and detector sandwiching an object to be inspected.

[0006] However, the steerable X-ray generators applied to such conventional X-ray inspection devices have the drawbacks of being complex in configuration and high in manufacturing costs, and the large-area detectors are also expensive components, which increases the unit price of the X-ray inspection device.

[0007] Another conventional X-ray inspection device applies a tomosynthesis-type cross-sectional image algorithm, which numerically combines multiple X-ray images while rotating an X-ray generator to obtain a cross-sectional image.

[0008] Tomosynthesis X-ray generators have non-steerable X-ray generators, which are cheaper than steerable X-ray generators for laminography, but have fewer functions. However, these X-ray generators have a wide beam angle, i.e., at least 100°, and are still expensive components compared to general X-ray generators.

[0009] To solve the above-mentioned problems, a low-cost CT (computer tomography) inspection device has been developed that scans parts of the object to be inspected, reconstructs the acquired image data into a tomographic image, and can determine defects using three-dimensional tomographic images.

[0010] Conventional CT inspection equipment takes multiple images of the electronic component being inspected from multiple angles, i.e., multiple projection images, and reconstructs these into a single CT image. However, the resolution and image quality of the final CT image are determined by the number of projection images.

[0011] Figure 1 shows an example of a CT tomographic image generated according to the number of projection images in a conventional CT examination device. As shown in Figure 1, it can be seen that the image quality and resolution of the CT tomographic image increase as the number of projection images increases.

[0012] Generally, a CT image generated using at least eight projection images is recognized as the minimum number of images required to determine whether a part is good or bad. However, it is known that even with a CT image generated using eight projection images, it is often impossible to determine the quality of some parts.

[0013] However, unlike medical images, in a CT inspection device for inspecting electronic components, it is practically impossible to capture a large number of projection images for each electronic component due to the resolution and image quality.

[0014] That is, if it takes a long time to inspect the quality of electronic components in an in-line process, it will cause delays in the entire manufacturing process, which will not only affect production volume but also increase the price of the product.

[0015] Therefore, technology that can generate CT images with high enough image quality and resolution to enable part quality determination even with a small number of projection images is recognized as an important element in actual production lines.

[0016] Furthermore, in relation to CT examination apparatuses, it would be preferable if a medical or dental CT examination apparatus could generate high-resolution, high-quality CT tomographic images by capturing a small number of projection images, as this would save examination time. [Prior art documents] [Patent documents]

[0017] [Patent Document 1] Korean Patent Registration No. 10-1717678 Summary of the Invention [Problem to be solved by the invention]

[0018] The present invention has been devised in view of the above-mentioned points, and its object is to provide a CT examination apparatus that can generate more accurate CT sectional images using a smaller number of projection images. [Means for solving the problem]

[0019] The above-mentioned object is achieved by the present invention by a CT inspection device comprising: a CT imaging unit that captures N projection images of an object to be inspected; and a CT image generation unit having an artificial intelligence-based AI model that receives the N projection images captured by the CT imaging unit and outputs a CT tomographic image of the object to be inspected.

[0020] Here, the AI ​​model is generated by learning using N training projection images, a first training CT tomographic image generated from the N training projection images, and a second training CT tomographic image generated from M (here, M>N) training projection images as training data, but can also be learned using the second training CT tomographic image as output data.

[0021] In addition, the AI ​​model is trained using the N training projection images and the first training CT tomographic image as input data, and the CT image generation unit may further include a preprocessing module that processes the N projection images, generates an input CT tomographic image, and inputs it into the AI ​​model.

[0022] Also, N can be 4 and M can be 8 or more. [Effects of the Invention]

[0023] According to the present invention, a CT examination apparatus is provided that can generate a CT sectional image with high image quality and resolution by capturing only a small number of projection images, for example, four projection images.

[0024] In addition, by capturing a small number of projection images, the inspection time is shortened, making it possible to perform CT inspections in real time in in-line processes, etc. [Brief explanation of the drawings]

[0025] [Figure 1]10A and 10B are diagrams showing examples of CT sectional images generated according to the number of projection images in a conventional CT examination device. [Figure 2] 1 is a diagram showing the configuration of a CT examination apparatus according to an embodiment of the present invention. [Figure 3] 1 is a diagram showing an example of imaging by a CT imaging unit of a CT examination apparatus according to an embodiment of the present invention. [Figure 4] FIG. 10 is a diagram showing an example of learning data applied to an AI model of a CT examination apparatus according to an embodiment of the present invention. [Figure 5] 3A to 3C are diagrams for explaining the results of CT tomographic image generation by the CT inspection apparatus according to the embodiment of the present invention. [Figure 6] 3A to 3C are diagrams for explaining the results of CT tomographic image generation by the CT inspection apparatus according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0026] The present invention relates to a CT inspection device, characterized in that it includes a CT imaging unit that captures N projection images of an object to be inspected, and a CT image generation unit that has an artificial intelligence-based AI model that receives input of the N projection images captured by the CT imaging unit and outputs a CT tomographic image of the object to be inspected. [Mode for carrying out the invention]

[0027] The advantages and features of the present invention, as well as methods for achieving them, will become apparent from the following detailed description of the embodiments in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, and can be realized in various forms. However, the present embodiments are provided to complete the disclosure of the present invention and to fully convey the scope of the present invention to those skilled in the art. The present invention is defined only by the scope of the claims.

[0028] The terms used in this specification are for the purpose of describing embodiments and are not intended to limit the present invention. In this specification, the singular forms "a," "an," and "the" include the plural forms unless the context clearly dictates otherwise. The terms "comprises" and / or "comprising" used in this specification do not exclude the presence or addition of one or more other elements in addition to the elements referenced. Throughout this specification, the same reference numerals refer to the same elements, and "and / or" includes each and every combination of one or more of the referenced elements. Although terms such as "first," "second," and the like are used to describe various elements, it goes without saying that these elements are not limited by these terms. These terms are merely used to distinguish one element from another. Therefore, a first element referenced below may be a second element within the technical spirit of the present invention.

[0029] Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used in a manner commonly understood by those skilled in the art to which the present invention pertains. Furthermore, terms defined in commonly used dictionaries are not to be interpreted in an ideal or excessive manner unless expressly defined otherwise.

[0030] The present invention will be described in detail below with reference to the accompanying drawings. In the embodiment of the present invention, a CT inspection device 100 will be described as being applied to the inspection of electronic components.

[0031] FIG. 2 is a diagram showing the configuration of a CT examination apparatus 100 according to an embodiment of the present invention.

[0032] 2, the CT examination apparatus 100 according to the embodiment of the present invention may include a CT imaging unit 110 and a CT image generating unit 120. The CT examination apparatus 100 according to the embodiment of the present invention may further include a control unit 130.

[0033] The CT imaging unit 110 according to an embodiment of the present invention captures N projection images of an electronic component. In the embodiment of the present invention, the CT imaging unit 110 captures four projection images, but when compared with M learning projection images used for learning an AI model 121 (described later), N is a natural number smaller than M, which will be described in detail later.

[0034] In the embodiment of the present invention, the CT imaging unit 110 captures four projection images at intervals of 90° as shown in FIG. 3, but the imaging direction is not limited to this.

[0035] On the other hand, the CT image generating unit 120 receives the N projection images captured by the CT imaging unit 110, for example, four projection images as described above, and generates a CT tomographic image of the electronic component.

[0036] In an embodiment of the present invention, the CT image generation unit 120 includes an artificial intelligence-based AI model 121, which receives input of N projection images and infers and outputs a high-quality CT tomographic image.

[0037] This makes it possible for the artificial intelligence-based AI model 121 to generate CT tomographic images with relatively high image quality and resolution using only four projection images, using a learning model that has already been learned.

[0038] Meanwhile, the AI ​​model 121 according to the embodiment of the present invention is generated by learning using, as learning data, N learning projection images, a first learning CT tomographic image generated from the N learning projection images, and a second learning CT tomographic image generated from M learning projection images, for example. Here, the first learning CT tomographic image and the second learning CT tomographic image are each learned as one image.

[0039] Here, the AI ​​model 121 is trained using the second training CT tomographic image as output data, and M is a natural number greater than N, as described above.

[0040] 4 is a diagram showing an example of training data applied to training of the AI ​​model 121 of the CT examination apparatus 100 according to the embodiment of the present invention. The training data can be stored in the training database 140.

[0041] As shown in Fig. 4(a), four learning projection images captured of a specific electronic component are input as a set of learning data. Then, the first learning CT tomographic image shown in Fig. 4(b) is a CT tomographic image generated using the four learning projection images shown in Fig. 4(a).

[0042] 4(c) is a CT tomographic image generated using M projection images captured of the electronic component. In the embodiment of the present invention, the second learning CT tomographic image used for learning is generated using eight or more projection images, and in order to improve the learning performance of the AI ​​model 121, a CT tomographic image generated using 360 projection images is used as the second learning CT tomographic image.

[0043] The AI ​​model 121 is trained using the training data described above, with the four training projection images and the first training CT tomographic image being used as input data, and the second training CT tomographic image being used as output data.

[0044] Here, the CT image generation unit 120 according to an embodiment of the present invention may include a pre-processing module 122 that processes N projection images captured by the CT imaging unit 110, generates one input CT tomographic image, and inputs it as input data for the AI ​​model 121.

[0045] That is, the control unit 130 controls the pre-processing module 122 to generate an input CT tomographic image using N projection images taken by the CT imaging unit 110, and the AI ​​model 121 receives the N projection images taken by the CT imaging unit 110 and the input CT tomographic image generated by the pre-processing module 122, infers the CT tomographic image, and outputs it.

[0046] 5 and 6 are diagrams for explaining examples of CT sectional images output by the CT inspection apparatus 100 according to the embodiment of the present invention.

[0047] In Figures 5 and 6, "Input 4P" indicates a CT tomographic image generated by processing four projection images taken by the CT imaging unit 110, "AI result" is the output of a CT tomographic image inferred by the AI ​​model 121 of the CT inspection device 100 according to an embodiment of the present invention, and "Correct answer 360P" is a CT tomographic image actually generated using 360 projection images, i.e., taken at 1° intervals.

[0048] As shown in Figures 5 and 6, it can be confirmed that the CT tomographic image output by the AI ​​model 121 according to an embodiment of the present invention is similar to the CT tomographic image generated using 360 actual projection images.

[0049] In other words, by capturing only four projection images, it is possible to generate CT images with high image quality and resolution, shortening inspection time and enabling the inspection of electronic components using CT images in real time during in-line processes.

[0050] In the embodiment of the present invention, the AI ​​model 121 is trained using a CNN-based Generative Adversarial Network (GAN).

[0051] In the above embodiment, the CT inspection apparatus 100 according to the embodiment of the present invention is described as being applied to the inspection of electronic components. However, the technical concept of the present invention is not limited to the inspection of electronic components, and can of course also be applied to CT inspection apparatuses for medical CT inspection and dental CT inspection.

[0052] This will enable the generation of CT images with higher image quality and resolution with fewer shots, even in medical CT examinations such as dental examinations, which is expected to reduce examination times.

[0053] Although the embodiments of the present invention have been described above with reference to the accompanying drawings, those skilled in the art will understand that the present invention can be embodied in other specific forms without changing the technical spirit or essential characteristics of the present invention. Therefore, it should be understood that the above-described embodiments are illustrative in all respects and are not limiting. [Industrial Applicability]

[0054] The technology can be applied to the field of inspection of small electronic components such as BGA (ball grid array) and CSP (chip scale package) on printed circuit boards (PCBs) that are incorporated as main components in electrical and electronic products such as home appliances and computers. [Explanation of symbols]

[0055] 100 CT inspection equipment 110 CT Imaging Department 120 CT image generation section 121 AI models 122 Pre-processing module 130 Control Unit 140 Learning DB

Claims

1. A CT examination device, a CT imaging unit that captures N projection images of the electronic component; a CT image generation unit having an AI model that has been trained to receive input of the N projection images captured by the CT imaging unit and output a CT tomographic image of the electronic component, The AI ​​model is N learning projection images, a first learning CT image generated from the N learning projection images, and a second learning CT image generated from M (here, M>N) learning projection images are generated by learning as learning data, learning is performed using the N learning projection images and the first learning CT image as input data; The second learning CT image is used as output data for learning; The CT inspection apparatus, characterized in that the CT image generation unit further includes a preprocessing module that processes N of the projection images, generates an input CT tomographic image, and inputs it to the AI ​​model.

2. 2. The CT examination apparatus according to claim 1, wherein N is 4 and M is 8 or more.

Images