JP7380342B2 - Line detection device and line detection program - Google Patents

Description

The present invention relates to a line segment detection device and a line segment detection program.

For fiber-reinforced resin molded products made of filler composite materials, such as carbon fiber-reinforced resins and glass fiber-reinforced resins, the fiber length greatly affects the mechanical properties, and the longer the fibers, the better the mechanical properties. It has been known. Therefore, in order to improve the performance and reliability of molded products, there is a need for evaluation techniques that can efficiently and accurately measure the fiber length in resin.

Patent Document 1 discloses that for a fiber image in which two or three fibers intersect, straight lines forming the outer shape of each fiber are found, and the intersection points of the straight lines are analytically determined from the equation of the obtained straight line. A technique has been disclosed in which the intersection points of fibers are determined by determining the intersection point, and the length of each fiber is calculated from the intersection point information.

Patent Document 2 discloses that for an image in which a plurality of fibers overlap, the intersection portions of the fibers are deleted from the image according to the thickness of the fibers and the angle of intersection, and the remaining fiber images are oriented in the same direction. A technique has been disclosed in which overlapping fibers are separated by reconnecting the fibers that overlap.

Patent Document 3 discloses that for an X-ray image of fiber-reinforced concrete, a representative value representing the state of the fibers is calculated based on an index correlated to the state of the fibers, and fiber parts can be identified from the X-ray image. After performing binarization processing and dividing the resulting binarized image into multiple regions, the fibers present in each divided region are recognized as line segments, and the dispersion state of the fibers is determined from the sum of the line segments. discloses a technique for determining the orientation state of fibers from the sum of sine length and cosine length.

JP2019-132686A Japanese Patent Application Publication No. 2002-259950 JP 2015-49099 Publication

The technology described in Patent Document 1 defines a straight line from the information on the outline of the fibers and finds the intersection point. Therefore, when a large number of fibers intersect, the amount of calculation required to determine the intersection point and calculate the intersection point coordinates is reduced. The number of fibers is enormous, and it takes time to measure the fiber length. Further, in the technique described in Patent Document 1, when fibers are detected as being divided, it is not possible to integrate the divided fibers and redefine them as a single fiber.

The technique described in Patent Document 2 intentionally removes fiber intersections from the image and reconnects them based on information on the remaining fibers, which may reduce the accuracy of detecting intersections. Further, in the technique described in Patent Document 2, the accuracy of reconnection may be lowered depending on the positional relationship of fibers, and the accuracy of measuring fiber length may be lowered.

The technology described in Patent Document 3 uses general image processing techniques such as strand processing that connects both end points and contour line processing that traces contour lines when recognizing fibers as line segments. If fibers or bent fibers are present, the ends of the fibers cannot be extracted. Therefore, the technique described in Patent Document 3 cannot correctly recognize one fiber as one line segment, and cannot accurately measure the fiber length.

The present invention has been made in view of the above points, and even when a plurality of fillers intersect in an image or a curved filler exists, it is possible to correctly detect each filler as an individual filler. The purpose of the present invention is to provide a line segment detection device and a line segment detection program.

A line segment detection device according to a first aspect of the present invention includes a binarization operation unit that binarizes an image taken of a composite material with a filler, and a binarization operation unit that converts the binarized image into a thin line. a thinning calculation unit that extracts feature points from the thinned image; a connected region detection calculation unit that searches for the feature points and detects a set of feature points to be connected; and a connected region detection calculation unit that searches for the feature points and detects a set of feature points to be connected. a line segment fitting calculation unit that evaluates continuity and linearity of the feature points and fits a plurality of line segments from the set of feature points detected by the line segment fitting calculation unit; and a line segment integration calculation unit that evaluates the continuity and linearity of line segments with respect to each other and integrates those that can be integrated into one line segment.

According to the first aspect of the present invention, the continuity and linearity of feature points are evaluated and a plurality of line segments are applied, and the continuity and linearity of the line segments are evaluated to determine what can be integrated. By integrating the line segments into one line segment, each line segment can be detected correctly even when a plurality of line segments intersect or there are curved line segments in the image.

A line segment detection device according to a second aspect of the present invention is the line segment detection device according to the first aspect of the present invention, which measures the length of each line segment in the image after being integrated by the line segment integration calculation unit. It further includes a line segment length measuring section.

According to the second aspect of the present invention, by measuring the length of each line segment in the image after being integrated by the line segment integration calculation unit, if a plurality of line segments intersect in the image or curved Even when line segments exist, the length of each line segment can be measured more accurately.

A line segment detection device according to a third aspect of the present invention is such that in the line segment detection device according to the first or second aspect of the present invention, the line segment fitting calculation unit is arranged to The shortest distance between a straight line connecting the starting point of the line segment and the candidate point and each feature point of the line segment is less than or equal to a predetermined threshold for all feature points of the line segment. In some cases, the candidate point is registered as a new feature point of the line segment.

According to the third aspect of the present invention, when a curved line segment exists in an image, what should be recognized as a line segment can be correctly recognized.

A line segment detection device according to a fourth aspect of the present invention is a line segment detection device according to the first or second aspect of the present invention, in which the line segment fitting calculation unit is configured to include a feature point adjacent to a feature point registered in the line segment immediately before. The feature point to be determined is set as a candidate point, and the angle formed by the vector of the line segment determined immediately before with respect to the vector consisting of the feature point registered in the line segment immediately before and the candidate point is less than or equal to a predetermined threshold. In this case, the candidate point is registered as a new feature point of the line segment.

According to the fourth aspect of the present invention, when a plurality of line segments intersect or curved line segments exist in an image, these can be recognized as a plurality of line segments.

A line segment detection device according to a fifth aspect of the present invention is a line segment detection device according to any one of the first to fourth aspects of the present invention, in which the line segment integration calculation section If the angle is less than a predetermined threshold and the distance between the end points of the two line segments is less than or equal to a predetermined threshold, the line segments are determined to be combinable and are combined into one line segment.

According to the fifth aspect of the present invention, even if line segments are separated, they can be recognized as one line segment if there is continuity.

A line segment detection device according to a sixth aspect of the present invention is a line segment detection device according to any one of the first to fourth aspects of the present invention, in which the line segment integration calculation section The number of feature points whose angle is less than or equal to a predetermined threshold and which exist between the two line segments is the value obtained by multiplying the distance between each end point between the two line segments by the predetermined threshold. If this is the case, it is determined that the line segment can be integrated, and the line segment is integrated into one line segment.

According to the sixth aspect of the present invention, even if line segments are separated, they can be recognized as one line segment if there is continuity.

A line segment detection device according to a seventh aspect of the present invention is the line segment detection device according to any one of the first to sixth aspects of the present invention, wherein the filler is an inorganic filler or an organic filler.

According to the seventh aspect of the present invention, the length of each filler can be accurately detected from an image taken of a material containing an inorganic filler or an organic filler.

A line segment detection program according to an eighth aspect of the present invention includes a computer including a binarization operation unit that binarizes an image taken of a composite material in which a filler is composited; a thinning calculation unit that thins an image and extracts feature points from the thinned image; a connected area detection calculation unit that searches for the feature points and detects a set of the feature points to be connected; and the connected area. a line segment fitting calculation unit that applies a plurality of line segments by evaluating the continuity and linearity of the feature points from the set of feature points detected by the detection calculation unit; Functions as a line segment detection device including a line segment integration calculation unit that evaluates the continuity and linearity of a plurality of line segments and integrates those that can be integrated into one line segment. .

According to the eighth aspect of the present invention, the continuity and linearity of feature points are evaluated and a plurality of line segments are applied, and the continuity and linearity of the line segments are evaluated to determine what can be integrated. By integrating the line segments into one line segment, each line segment can be detected correctly even when a plurality of line segments intersect or there are curved line segments in the image.

According to the present invention, there is provided a line segment detection device that can correctly detect each filler as a line segment even when a plurality of fillers intersect or a curved filler exists in an image, and a line segment detection device that can correctly detect each filler as a line segment. A minute detection program can be provided.

FIG. 1 is a block diagram showing the hardware configuration of a line segment detection device according to the present embodiment. FIG. 2 is a block diagram showing an example of a functional configuration of a line segment detection device. 3 is a flowchart showing the flow of line segment length measurement processing performed by the line segment detection device. 3 is a flowchart illustrating line segment fitting processing. It is a figure explaining line segment fitting processing. It is a figure explaining line segment fitting processing. 3 is a flowchart illustrating line segment integration processing. It is a figure explaining line segment integration processing. 3 is a flowchart illustrating line segment integration processing. FIG. 2 is a diagram showing an example of an image input to a line segment detection device. It is a figure which shows the example of the line segment detected by the line segment detection apparatus. It is a graph explaining the effect of a line segment detection device.

An example of an embodiment of the present disclosure will be described below with reference to the drawings. In addition, the same reference numerals are given to the same or equivalent components and parts in each drawing. Furthermore, the dimensional ratios in the drawings are exaggerated for convenience of explanation and may differ from the actual ratios.

FIG. 1 is a block diagram showing the hardware configuration of a line segment detection device 10 according to this embodiment. The line segment detection device 10 is a device that detects a line segment, that is, a filler, in an image of a composite material containing a filler using a microscope or the like, and calculates the length of the line segment. .

As shown in FIG. 1, the line segment detection device 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage 14, an input unit 15, a display unit 16, and a communication unit. It has an interface (I/F) 17. Each configuration is communicably connected to each other via a bus 19.

The CPU 11 is a central processing unit that executes various programs and controls various parts. That is, the CPU 11 reads a program from the ROM 12 or the storage 14 and executes the program using the RAM 13 as a work area. The CPU 11 controls each of the above components and performs various arithmetic operations according to programs recorded in the ROM 12 or the storage 14. In this embodiment, the ROM 12 or the storage 14 stores a line segment detection program that detects line segments from an image of a material composited with filler and measures the length of the detected line segment.

The ROM 12 stores various programs and data. The RAM 13 temporarily stores programs or data as a work area. The storage 14 is constituted by a storage device such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or a flash memory, and stores various programs including an operating system and various data.

The input unit 15 includes a pointing device such as a mouse and a keyboard, and is used to perform various inputs.

The display unit 16 is, for example, a liquid crystal display, and displays various information. The display section 16 may employ a touch panel method and function as the input section 15.

The communication interface 17 is an interface for communicating with other devices, and uses standards such as Ethernet (registered trademark), FDDI, and Wi-Fi (registered trademark), for example.

When executing the above line segment detection program, the line segment detection device 10 uses the above hardware resources to realize various functions. The functional configuration realized by the line segment detection device 10 will be explained.

FIG. 2 is a block diagram showing an example of the functional configuration of the line segment detection device 10.

As shown in FIG. 2, the line segment detection device 10 includes a binarization calculation section 101, a line thinning calculation section 102, a connected area detection calculation section 103, a line segment fitting calculation section 104, and a line segment integration calculation section. 105, and a line segment length measuring section 106. Each functional configuration is realized by the CPU 11 reading out and executing a line segment detection program stored in the ROM 12 or the storage 14.

The binarization calculation unit 101 executes a process of binarizing an image of a composite material containing a filler, which is input to the line segment detection device 10 . In this embodiment, the binarization calculation unit 101 executes binarization processing on a fiber image obtained by capturing a composite material in which metal fibers, carbon fibers, or glass fibers are combined as fillers, for example. Binarization is a process for separating composite fibers and background from a fiber image. Known processing such as adaptive threshold processing is used for binarization.

The thinning calculation unit 102 performs thinning processing on the image binarized by the binarization calculation unit 101, and extracts feature points from the thinned image. Thinning is a process for detecting the center line of fibers in a binarized image. For thinning, a known process such as the Zhang-Suen thinning algorithm is used.

The connected region detection calculation unit 103 searches for feature points extracted from the thinned image and detects a set of feature points to be connected. The connected region detection calculation unit 103 labels each connected feature point group. By labeling each connected feature point group by the connected region detection calculation unit 103, the line segment detection device 10 can perform subsequent processing for each labeled point group, thereby improving calculation efficiency.

The line segment fitting calculation unit 104 evaluates the continuity and linearity of the feature points from the set of feature points (feature point group) detected by the connected area detection calculation unit 103, and fits a line segment. Details of the line segment fitting process to the feature point group will be described in detail later.

The line segment integration calculation unit 105 evaluates the continuity and linearity of the line segments fitted by the line segment fitting calculation unit 104, and integrates those that can be integrated into one line segment. do. When multiple fibers overlap or when the fibers are curved and not straight, multiple line segments may be divided and applied to one fiber. Therefore, the line segment integration calculation unit 105 detects the set of line segments fitted by the line segment fitting calculation unit 104 from the continuity and linearity of the line segments, and detects those that should be one line segment. Combine multiple line segments into one line segment. Details of the line segment integration process will be described later.

The line segment length measurement unit 106 measures the length of each line segment in the image after the line segment integration calculation unit 105 has integrated it. Information on the length of the line segment measured by the line segment length measurement unit 106 is output from the line segment detection device 10.

Next, the operation of the line segment detection device 10 will be explained.

FIG. 3 is a flowchart showing the flow of line segment length measurement processing by the line segment detection device 10. The line segment length measurement process is performed by the CPU 11 reading a line segment detection program from the ROM 12 or storage 14, loading it into the RAM 13, and executing it.

The CPU 11 first executes a fiber image input process in which a fiber image obtained by imaging a composite material containing an inorganic filler or an organic filler as a filler is input into the line segment detection device 10 (step S11). The type of inorganic filler is not particularly limited. Specifically, silicates (silicate minerals) such as talc, silica, clay, montmorillonite or kaolin, carbonates such as calcium carbonate, lithium carbonate or magnesium carbonate, hydroxides such as aluminum hydroxide or magnesium hydroxide, Sulfides such as barium sulfate, metal oxides such as alumina, titanium oxide or zinc oxide, titanides such as potassium titanate or barium titanate, metals such as aluminum, iron, silver or copper, glass powder or glass fiber, etc. Examples include glasses, carbides such as charcoal or bamboo charcoal, carbons such as fullerene, carbon nanotubes, and carbon fibers. These inorganic fillers may be natural or artificial. Moreover, only one type of these inorganic fillers may be used, or two or more types may be used in combination. The type of organic filler is not particularly limited. Examples of organic fibers include aromatic polyester fibers, aromatic polyamide fibers, fluororesin fibers, polyimide fibers, vegetable fibers, and the like. These organic fillers may be natural or artificial. Moreover, only one type of these organic fillers may be used, or two or more types may be used in combination.

Following step S11, the CPU 11 executes binarization processing on the input fiber image (step S12). This binarization process is a process for separating the composite fibers and the background from the fiber image as described above. The CPU 11 executes binarization processing using known processing such as adaptive threshold processing.

Following step S12, the CPU 11 executes thinning processing on the fiber image after the binarization processing (step S13). As described above, this thinning process is a process for detecting the center line of the fiber in the binarized image. The CPU 11 executes line thinning processing using known processing such as the Zhang-Suen line thinning algorithm.

Following step S13, the CPU 11 searches for feature points extracted from the thinned image and executes a connected region extraction process to detect a set of feature points to be connected (step S14).

Following step S14, the CPU 11 executes a line segment fitting process that evaluates the continuity and linearity of the feature points and fits a plurality of line segments from the set of feature points detected in step S14 (step S15). . Details of the line segment fitting process will be described later.

Following step S15, the CPU 11 evaluates the continuity and linearity of the plurality of line segments fitted in step S15, and selects line segments that can be integrated into one line segment. Integration processing is executed (step S16). Details of the line segment integration process will be described later.

Following step S16, the CPU 11 executes line segment length calculation processing to calculate the length of the line segment in the fiber image after performing the line segment integration processing in step S16 (step S17).

Following step S17, the CPU 11 outputs the coordinates and the total length of the line segment calculated in step S17 (step S18).

By executing the series of processes shown in FIG. 3, the CPU 11 can correctly detect individual line segments even when multiple line segments intersect or curved line segments exist in the image. Then, the length of the detected line segment can be calculated.

Next, details of the line segment fitting process in step S15 in FIG. 3 will be explained.

FIG. 4 is a flowchart illustrating the line segment fitting process in step S15 in FIG. Further, FIGS. 5A and 5B are diagrams illustrating line segment fitting processing.

The CPU 11 first sorts the labeled feature points extracted from the thinned image by coordinates (step S101). When sorting the feature points by coordinates, the CPU 11 may, for example, set the upper left of the thinned image as the origin.

Following step S101, the CPU 11 determines whether there are any remaining feature points (step S102). If it is determined that there are no remaining feature points (step S102; No), the CPU 11 ends the series of line segment fitting processes. Following step S102, if it is determined that there are remaining feature points (step S102; Yes), the CPU 11 sets the first feature point after sorting, for example, the upper left feature point, as the starting point ( Step S103). In the example shown in FIG. 5A, the CPU 11 sets the first feature point located at the upper left as the starting point.

Following step S103, the CPU 11 selects candidate points from points adjacent to the end point of the line segment (step S104).

Following step S104, the CPU 11 determines whether there are candidate points, that is, whether the candidate points were selected in step S104 (step S105). If it is determined that the candidate point has been selected (step S105; Yes), the CPU 11 then determines whether the candidate point selected in step S104 can be added to the line segment (step S106). If it is determined that the candidate point can be added to the line segment (step S106; Yes), the CPU 11 adds the candidate point selected in step S104 to the line segment (step S107). After adding the candidate point to the line segment, the CPU 11 returns to the process of step S104.

If it is determined that the candidate point cannot be selected as a result of the determination in step S105 (step S105; No), or if it is determined that the candidate point cannot be added to the line segment as a result of the determination in step S106 (step S106; ;No), the CPU 11 registers the line segments generated from the feature points in the processing up to this point, and removes the feature points that constitute the registered line segments from the feature point group (step S108). After registering the line segment and removing the registered line segment feature points from the feature point group, the CPU 11 returns to step S102, and if there are remaining feature points, selects the first feature point from the remaining feature point group, For example, the upper leftmost feature point is set as the starting point.

In this way, the CPU 11 repeats the search for feature points as long as the adjacent points satisfy the linearity condition. If linearity is no longer satisfied or if there is no adjacent point to the last added point, the CPU 11 ends the search and determines a line segment from the starting point and ending point of the points traced up to that point. The CPU 11 detects a plurality of line segments by repeating the same process on the remaining feature points.

Using the example shown in FIG. 5A, the CPU 11 first sets the number 1 feature point as the starting point, and then adds the number 1 feature point and the number 2 feature point adjacent to the lower right to the line segment. Then, the second feature point and the third feature point adjacent to the lower right are added to the line segment. Then, the CPU 11 adds the number 3 feature point and the number 4 feature point adjacent to the lower right to the line segment.

If feature point No. 4 is added, feature points No. 5 and 6 will be adjacent to feature point No. 4. Let a be a vector directed from the first feature point, which is the starting point, to the fourth feature point registered in the line segment immediately before. On the other hand, feature points No. 5 and 6 adjacent to feature point No. 4 are candidate points, and vectors directed toward feature points No. 5 and 6 with reference to feature point No. 4 are designated as b and c, respectively. If there are two or more candidate points, the CPU 11 narrows down the candidate points to one based on the direction of the candidate points with respect to the line segments registered so far. The angle θ _ab between vectors a and b and the angle θ _ac between vectors a and c are determined by the following equation (1).

If the critical angle that satisfies linearity is defined as θ _thr , then when θ _ab ≦θ _thr or θ _ac ≦θ _thr is satisfied, the CPU 11 sets the 5th or 6th feature point to a new feature point of the line segment. Register as a candidate point. However, if two or more candidate points satisfy the above conditions, the CPU 11 selects a point that satisfies linearity more among the candidate points, that is, a feature point that has the smallest angle with vector a, as a line segment. is registered as a new feature point candidate point.

After narrowing down the candidate points of the line segment, the CPU 11 determines whether the candidate point is a feature point of the line segment based on the linearity of the line segment. A method for determining linearity will be described using as an example the positional relationship between feature points and candidate points shown in FIG. 5B. First, the CPU 11 defines a straight line connecting the starting point and the candidate point. The broken line in FIG. 5B is a straight line connecting the starting point and the candidate point. Then, the CPU 11 calculates the shortest distance l _d from the defined straight line for all the feature points already registered as line segments, and if all of them are within a certain predetermined threshold, the CPU 11 moves the candidate point to the line. Register as a new feature point.

Here, it is assumed that the CPU 11 registers feature point No. 5 as a new feature point of the line segment. In this case, vector a is a vector directed from feature point No. 1, which is the starting point, to feature point No. 5. Then, the CPU 11 registers the feature point No. 7 adjacent to the feature point No. 5 and the feature point No. 9 adjacent to the feature point No. 7 as line segments.

Next, details of the line segment integration process in step S16 in FIG. 3 will be described.

FIG. 6 is a flowchart illustrating the line segment integration process in step S16 of FIG. Further, FIG. 7 is a diagram illustrating line segment integration processing. In FIG. 7, let d and e be vectors of line segments that are integration candidates, and let f be a vector connecting the end points of vectors d and e, that is, a vector from the end point of vector d to the start point of vector e.

The CPU 11 first determines whether the angle θ _de between the vector d and the vector e of the line segments of the integration candidates is less than or equal to the critical angle θ _thr1 that satisfies linearity, that is, whether θ _de ≦θ _thr1 is satisfied. Determination is made (step S111).

As a result of the determination in step S111, if θ _de ≦θ _thr1 is satisfied (step S111; Yes), the CPU 11 then determines whether the size |f| of the vector f is less than or equal to the critical length l _thr for integration determination. That is, it is determined whether |f|≦l _thr is satisfied (step S112). On the other hand, as a result of the determination in step S111, if θ _de ≦θ _thr1 is not satisfied (step S111; No), the CPU 11 determines that the two line segments cannot be integrated (step S117).

As a result of the determination in step S112, if |f|≦l _thr is satisfied (step S112; Yes), the CPU 11 determines that the two line segments can be integrated (step S116). On the other hand, as a result of the determination in step S112, if |f|≦l _thr is not satisfied (step S112; No), then the CPU 11 sets a predetermined ratio r ₁ ( 0<r ₁ ≦1) and the size of the vector f |f| It is determined whether |)×r ₁ is satisfied (step S113).

As a result of the determination in step S113, if it is determined that |f|>min(|d|, |e|)×r ₁ is satisfied (step S113; Yes), the CPU 11 determines that the two line segments cannot be integrated. (Step S117). On the other hand, as a result of the determination in step S113, if it is determined that |f|>min(|d|, |e|)×r ₁ is not satisfied (step S113; No), then the CPU 11 It is determined whether two line segments can be integrated based on the proportion of feature points. That is, the CPU 11 calculates the shortest distance from the vector f for each of the feature points that are close to the vector f, and determines that the sum of lengths (number of pixels) l _online of the feature points that exist within a certain predetermined threshold is l _online ≦ It is determined whether |f|×r ₂ (0<r ₂ ≦1) is satisfied (step S114). In the example of FIG. 7, the number of feature points for which the shortest distance l _f between the vector f and the feature point is less than or equal to a predetermined threshold is l _online .

As a result of the determination in step S114, if it is determined _that l _online ≦|f| It is determined that integration is not possible (step S117). On the other hand, as a result of the determination in step S114, if it is determined that l _online ≦|f _| It is determined that the two line segments can be integrated (step S116).

In this way, the CPU 11 can evaluate the continuity and linearity of two line segments, and integrate those that can be integrated into one line segment.

When determining whether two line segments can be integrated, the CPU 11 further determines that the angle θ _df between the vector d and the vector f, and the angle θ _ef between the vector e and the vector f are both equal to or less than a threshold value. It may be determined whether or not the two line segments can be integrated depending on whether or not the two line segments can be integrated.

FIG. 8 is a flowchart illustrating the line segment integration process in step S16 in FIG. The flowchart shown in FIG. 8 has one process added after step S114 in the flowchart shown in FIG.

As a result of the _{determination} in step S114, if it is determined _that l _online ≦|f| It is determined whether the angle θ _ef formed by the vector f and the vector f are both less than or equal to the threshold value θ _thr2 . That is, the CPU 11 determines whether θ _df ≦θ _thr2 and θ _ef ≦θ _thr2 are satisfied (step S115).

As a result of the determination in step S115, if θ _df ≦θ _thr2 and θ _ef ≦θ _thr2 are satisfied (step S115; Yes), the CPU 11 determines that the two line segments can be integrated (step S116). On the other hand, as a result of the determination in step S115, if either θ _df ≦θ _thr2 or θ _ef ≦θ _thr2 is not satisfied (step S115; No), the CPU 11 determines that the two line segments cannot be integrated ( Step S117).

The effects of the line segment detection device 10 according to the present embodiment will be described with reference to a dispersed image of carbon fiber (CF).

FIG. 9 is a diagram showing an example of a fiber dispersion image. The image shown in Figure 9 was obtained by heating an injection molded product of carbon fiber reinforced plastic (CFRP), extracting only the carbon fibers, dispersing and fixing them in a petri dish, and imaging them with a microscope. This is the image that was created. The black part in the image is carbon fiber, and there are many fibers with different lengths.

FIG. 10 is a diagram showing line segments detected by the line segment detection device 10 according to this embodiment. In FIG. 10, line segments detected by the line segment detection device 10 are shown by thick lines. As shown in FIG. 10, it can be seen that the line segment detection device 10 correctly detects crossed fibers and long fibers.

In order to verify accuracy, the weight average fiber length L _w was investigated based on the line segments detected by the line segment detection device 10 and compared with the results of manual measurement. The weight average fiber length L _w is the average value of fiber lengths weighted by length, and is calculated using the following formula (2).

In formula (2), l _i is the fiber length, and N is the number of measured fibers.

In manual measurement, first, noise removal, color tone correction, and preprocessing for binarization were performed manually using commercially available image processing software, and then the fiber length was automatically measured using the automatic measurement function of the same software. Among the results, the poorly detected fiber data were manually corrected and used as data for comparison.

FIG. 11 is a graph showing the comparison results between manual measurement and this embodiment. In this verification, fibers with a diameter of 0.05 mm or less were excluded from the count. In addition, the total number of fibers in this verification was approximately 4,000.

As shown in FIG. 11, the difference between the measured value by the line detection device 10 according to this embodiment and the manually measured value was 5% or less. Therefore, it can be seen that the line segment detection device 10 according to this embodiment can measure the fiber length with high accuracy.

The line segment detection device 10 according to the present embodiment evaluates the linearity and continuity of the line segments even when a plurality of line segments intersect in the image or when there are curved line segments, By integrating line segments that can be integrated, individual line segments can be detected correctly.

Note that the line segment length measurement process executed by the CPU reading the software (program) in each of the above embodiments may be executed by various processors other than the CPU. In this case, the processor includes a PLD (Programmable Logic Device) whose circuit configuration can be changed after manufacturing, such as an FPGA (Field-Programmable Gate Array), and an ASIC (Application Specific Integrated Cipher). rcuit) to execute specific processing such as An example is a dedicated electric circuit that is a processor having a specially designed circuit configuration. In addition, the line segment length measurement process may be executed by one of these various processors, or by a combination of two or more processors of the same type or different types (for example, multiple FPGAs, and a combination of a CPU and an FPGA). combinations etc.). Further, the hardware structure of these various processors is, more specifically, an electric circuit that is a combination of circuit elements such as semiconductor elements.

Further, in each of the above embodiments, a mode has been described in which a program for line segment length measurement processing is stored (installed) in advance in the ROM or storage, but the present invention is not limited to this. The program can be stored in non-temporary memory such as CD-ROM (Compact Disk Read Only Memory), DVD-ROM (Digital Versatile Disk Read Only Memory), and USB (Universal Serial Bus) memory. (non-transitory) recorded on a recording medium It may be provided in the form of Further, the program may be downloaded from an external device via a network.

10 Line segment detection device 101 Binarization calculation unit 102 Line thinning calculation unit 103 Connected region detection calculation unit 104 Line segment fitting calculation unit 105 Line segment integration calculation unit 106 Line segment length measurement unit

Claims (8)

a binarization calculation unit that binarizes an image taken of a composite material containing a filler;
a thinning calculation unit that thins the image binarized by the binarization calculation unit and extracts feature points from the thinned image;
a connected region detection calculation unit that searches for the feature points and detects a set of the feature points to be connected;
a line segment fitting calculation unit that evaluates continuity and linearity of the feature points and fits a plurality of line segments from the set of feature points detected by the connected region detection calculation unit;
a line segment integration calculation unit that evaluates the continuity and linearity of the line segments fitted by the line segment fitting calculation unit and integrates those that can be integrated into one line segment; ,
A line segment detection device comprising: The line segment detection device according to claim 1, further comprising a line segment length measuring unit that measures the length of each line segment in the image after being integrated by the line segment integration calculation unit. The line segment fitting calculation unit takes a feature point adjacent to the feature point registered in the line segment immediately before as a candidate point, and calculates a straight line connecting the starting point of the line segment and the candidate point, and each feature point of the line segment. The line segment according to claim 1 or 2, wherein the candidate point is registered as a new feature point of the line segment when the shortest distance between all feature points of the line segment is less than or equal to a predetermined threshold value for all feature points of the line segment. Detection device. The line segment fitting calculation unit sets a feature point adjacent to a feature point registered in the line segment immediately before as a candidate point, and calculates, for a vector consisting of the feature point registered in the line segment immediately before and the candidate point, Line segment detection according to claim 1 or 2, wherein the candidate point is registered as a new feature point of the line segment when the angle formed by the vector of the line segment determined just before is less than or equal to a predetermined threshold. Device. The line segment integration calculation unit is configured to perform integration when the angle formed by the vectors of the two line segments is less than or equal to a predetermined threshold, and the distance between the end points of the two line segments is less than or equal to a predetermined threshold. The line segment detection device according to any one of claims 1 to 4, which determines that the lines are line segments and integrates them into one line segment. The line segment integration calculation unit calculates the angle between the two line segment vectors to be less than or equal to a predetermined threshold, and the number of feature points existing between the two line segments to be calculated by the line segment integration calculation unit. Any one of claims 1 to 4, wherein if the distance between each end point between the lines is equal to or greater than a value multiplied by a predetermined threshold value, the line segment is determined to be combinable and is integrated into one line segment. The line segment detection device according to item 1. The line detection device according to claim 1, wherein the filler is an inorganic filler or an organic filler. computer,
a binarization calculation unit that binarizes an image taken of a composite material containing a filler;
a thinning calculation unit that thins the image binarized by the binarization calculation unit and extracts feature points from the thinned image;
a connected region detection calculation unit that searches for the feature points and detects a set of the feature points to be connected;
a line segment fitting calculation unit that evaluates continuity and linearity of the feature points and fits a plurality of line segments from the set of feature points detected by the connected region detection calculation unit;
a line segment integration calculation unit that evaluates the continuity and linearity of the line segments fitted by the line segment fitting calculation unit and integrates those that can be integrated into one line segment; ,
A line segment detection program for functioning as a line segment detection device equipped with.