JPH09329659A - Target discrimination device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a target discrimination device capable of accurately discriminating a target. SOLUTION: This device is provided with a preprocessing part 12 cutting out the shape of target from digital image data coming from an image input part 11 and extracting the photographing shape data, a three dimensional data memory 13 storing the three dimensional data of target candidates, a target cross section calculation part 14 calculating the azimuth of the target position from the relative position relation between a camera at the photographing time obtained via a photographing situation data input line 17 and the photographed target and calculating a cross section shape of the observed target from the azimuth angle based on the three dimensional data, and a comparison and discrimination part 15 discriminating the target by comparing the photographed shape data of the target from the preprocessing part 12 with the cross section shape calculated by the target cross section calculation part 14. By comparing the target silhouette for the case the target is seen from the position rotated for the depression angle of the camera, and the target silhouette actually photographed, the target is identified.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a target discriminating apparatus for discriminating a target shown in an image taken by an image pickup means such as an optical camera and a video camera.

[0002]

2. Description of the Related Art Conventionally, various devices have been provided for recognizing a target, such as a character or an object included in image pickup data, by matching the shape of the target with the original data. For example, FIG. 11 is a block diagram of a conventional character recognition device for image recognition shown in "Technical Research Report of Institute of Electronics, Information and Communication Engineers" (Vol. 92 No. 433, January 21, 1993). The apparatus shown in FIG. 11 includes a character cutout unit 1 for inputting a character from an image input line 6, a preprocessing unit 2 for performing processing such as noise removal, and a feature extraction unit 3 for extracting a feature amount of the cutout character. And a matching pattern storage unit 4 that stores character shape information that is original data of characters,
The matching pattern storage unit 4 identifies a character based on the character shape information and outputs the result from the determination result output line 7.

Next, the operation of this apparatus will be described.

Characters to be recognized are partitioned and input for each character by the character slicing unit 1 and sent to the preprocessing unit 2. The pre-processing unit 2 performs noise removal of input characters, normalization of size, etc., and the feature extraction unit 3
Send to The feature extraction unit 3 extracts the contour line of the preprocessed character, counts the direction index of each contour line, and sends the calculated feature amount of the character to the identification unit 5. The identification unit 5 calculates the characteristic amount of the sent character by an identification function to make a determination, selects the closest character, and outputs it as another result.

A target identifying device for identifying a moving target object is disclosed in Japanese Patent Application Laid-Open No. 7-174837.

[0006]

Since the conventional target discriminating apparatus is constructed as described above, the target to be discriminated to be discriminated in the image must be clearly projected without blurring, and from a distance. The accuracy in determining the shooting target was not good. In addition, since the outer shape of the target is used for the determination, the shape may change depending on the angle at which the target is imaged, and further, it becomes extremely difficult to perform the determination for the targets having similar shapes. There was a point.

The present invention has been made to solve the above problems, and its object is to blur an image due to a change in outer shape depending on the apparent azimuth of the target, out-of-focus of the image pickup means, or photographing from a distance. An object of the present invention is to provide a target discriminating apparatus capable of discriminating a target with high accuracy even when the target has been exhausted.

It is another object of the present invention to provide a target discriminating device which can discriminate between targets having similar outer shapes as candidates for the photographed targets.

[0009]

In order to achieve the above-mentioned object, a target discriminating apparatus according to the present invention is a target discriminating apparatus for discriminating a target displayed in an image photographed by an image pickup means. Target shape data extraction means for obtaining imaged shape data of the imaged target, target shape data storage means for storing three-dimensional data of target candidates,
The azimuth at which the target is located is calculated from the positional relationship between the image pickup means and the target at the time of shooting, and the cross-sectional shape when the target is viewed from the azimuth angle is calculated based on the three-dimensional data stored in the target shape data storage means. A target cross-section calculation means, a target imaging shape data extracted by the target shape data extraction means, and a cross-sectional shape calculated by the target cross-section calculation means. It is characterized by having.

Further, the comparison / discrimination means is included in the specific shape included in the cross-sectional shape calculated by the target cross-section calculating means and the target imaging shape data extracted by the target data extracting means corresponding to the specific shape. The characteristic shape is further compared. Furthermore, the peculiar shape is characterized in that it is a protrusion.

According to another aspect of the present invention, in a target discriminating apparatus for discriminating a target displayed in an image photographed by the image pickup means, at the same time as the image pickup means, assistance data detection means for detecting assistance data relating to the same target; Target shape data extraction means for obtaining the imaged shape data of the target imaged by means, support information generation means for generating target discrimination support information based on the support data by the support data detection means, shape data of target candidates and target candidates Target information storage means for storing target specifying information relating to the characteristic items of
A comparison determination unit that determines a target based on a comparison result between the target imaging shape data and the target candidate shape data and a comparison result between the target determination support information and the target specifying information; It is characterized in that the target is discriminated based on the imaging shape data and the target support information.

Further, in the above-mentioned other invention, the support information generation means obtains the heat source position of the target extracted based on the infrared image of the target by the support data detection means as the target discrimination support information. Means, the target information storage means stores the heat source position of the target candidate as target specifying information, and the comparison and determination means, the target heat source position generated by the support information generation means, and the target information storage means. Is compared with the heat source position of the target candidate stored by.

Further, in the above-mentioned other invention, the support information generation means uses the target dimension as the target determination support information from the distance measurement data to the target obtained based on the imaging data obtained by copying the target by the support data detection means. A target dimension calculation unit for calculating, the target information storage unit stores the target candidate size information as target specifying information, and the comparison and determination unit, the target size calculated by the support information generation unit; It is characterized in that the size information of the target candidate stored in the target information storage means is compared.

Further, in the above-mentioned other invention, the support information generating means includes a geographic data storage section for storing geographic information and coordinate information, image pickup data showing a target by the support data detecting means and the geographic data storage. A topographical information determination means for generating target position information obtained as a result of comparison with the information stored in the division as target determination support information, and the target information storage means uses topographical information regarding target candidates as target identification information. The comparison determination means stores the target position information calculated by the support information generation means and the target candidate topographical information stored in the target information storage means.

Further, in the above-mentioned other invention, the support information generating means is detected by an electronic equipment specification data storage section for storing specification data of an electronic equipment which can be mounted on the target, and the support data detecting means. A radio wave data radiated by a target and a mounted electronic device determination unit that generates a target mounted electronic device identified based on the specification data stored in the electronic device specification data storage unit as target determination support information, The target information storage means stores an electronic device mounted on a target candidate as target specifying information, and the comparison and determination means stores the target mounted electronic device calculated by the support information generation means and the target information storage means. It is characterized in that it is compared with an electronic device of a target candidate to be set.

Further, in the target discriminating apparatus for discriminating the target displayed in the image photographed by the image pickup means, the target shape data extracting means for obtaining the image pickup shape data of the target photographed by the image pickup means and the speed performance of each target candidate. Speed performance specification data storage means for storing information, speed data for the same target detected at the same time as the imaging means, and target candidates that can be determined by comparison with speed performance information stored in the speed performance specification data storage means. Target speed performance selecting means for selecting the target speed performance selecting means, target shape data storing means for storing shape data of target candidates, and target candidates selected by the target speed performance selecting means from the shape data stored in the target shape data storing means. Shape data for the target imaging shape data extracted by the target shape data extracting means,
And a comparison / determination unit that determines the target by comparing

Further, in a target discriminating device for discriminating a target displayed in an image photographed by the image pickup means, a target information storage means for storing shape data of the target candidate and formation formation information regarding the formation, formation etc. taken by the target candidate. , Target formation data storage means for storing formation formation information such as formations and formations that the target candidate can take; target shape data extraction means for obtaining each image pickup shape data of a plurality of targets imaged by the image pickup means; Target formation detecting means for obtaining the arrangement situation of a plurality of targets imaged by means, and formation formation information corresponding to the arrangement situation obtained by the target arrangement detecting means of formation formation information stored in the target arrangement data storage means. A target placement determining unit that identifies the target placement shape, a comparison result of the target imaging shape data and the target candidate shape data, and the target placement determination Means and having a comparison discriminating means for discriminating a target based on a result of comparison between the identified formation arrangement information and the formation location information target information storage means is stored.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of the present invention will be described below with reference to the drawings. Note that, in each embodiment, the same reference numerals are given to the same elements, and the description thereof will be omitted when they overlap.

Embodiment 1 FIG. 1 is a block diagram showing a first embodiment of a target discriminating apparatus according to the present invention. The apparatus shown in FIG. 1 includes an image input unit 11, a preprocessing unit 12, a three-dimensional data storage unit 13, a target cross-section calculation unit 14, and a comparison determination unit 15. The image input unit 11
It is an image data input means for inputting target image pickup data photographed by an image pickup means (not shown) such as an optical camera or a video camera as digital data through the image input line 16. The preprocessing unit 12 is a target data extraction unit that cuts out a target region, that is, a target shape from the input image, and extracts target imaging shape data, and forms a target shape data extraction unit together with the image input unit 11. The three-dimensional data storage unit 13 is a target shape data storage unit that stores three-dimensional data of target candidates. The target cross-section calculation unit 14 calculates the azimuth at which the target is located from the data representing the relative positional relationship between the imaging means and the imaged target at the time of shooting, which is obtained through the shooting situation data input line 17, and from the azimuth angle. It is a target cross-section calculation means for calculating the cross-sectional shape when the target is viewed based on the three-dimensional data. The comparison determination unit 15 is a comparison determination unit that determines the target by comparing the target imaging shape data from the preprocessing unit 12 and the cross-sectional shape calculated by the target cross-section calculation unit 14, and determines the determination result. The result is output from the output line 18.

A feature of this embodiment is that the target cross-sectional shape, that is, the silhouette is used as information for comparing the target shapes. That is, since the silhouette of the imaged target is compared with the silhouette of the three-dimensional data that is the original data of the target when viewed from the position rotated based on the depression angle of the camera that captured the target, Even when the shape of the target looks different depending on the viewing angle, it is possible to correctly determine what the target is.

Next, the operation of this embodiment will be described.

First, a video image of a target taken by a camera which is an image pickup means is input to the image input section 11 from the image input line 16. In the image input unit 11, the input image is converted into digital data and the pre-processing unit 12 converts it into one image.
To be transmitted. The pre-processing unit 12 performs processing for cutting out a target area to be discriminated, that is, a target shape from the digital image. Here, general image processing such as density conversion and noise removal is performed, and then the target target area is cut out automatically or manually. The cut-out area is sent to the comparison / determination unit 15. This extraction process may be performed using a conventional method.

On the other hand, data representing the relative positional relationship between the camera and the target when the target is photographed is obtained by the photographing condition data input line 1
It inputs from 7 to the target section calculation part 14. The target cross-section calculation unit 14 calculates the cross-sectional shape of the target by rotating the three-dimensional data of each target candidate stored in the three-dimensional data storage unit 13 based on the azimuth angle that is the input shooting situation data. To do. In this way, the cross-sectional shape, that is, the matching pattern, corresponding to the target outer shape that changes depending on the angle at which the target is photographed is calculated. The calculated matching pattern is transmitted to the comparison / determination unit 15.

In the comparison / discrimination unit 15, the imaging shape data obtained based on the actually imaged target sent from the preprocessing unit 12 and the matching pattern obtained based on the original data sent from the target section calculation unit 14 And a target candidate of a matching pattern whose shape can be determined to be the closest is output from the determination result output line 18 as a determination result.

As described above, according to the present embodiment, the silhouette of the target actually captured is compared with the silhouette of the target candidate when viewed from the position rotated based on the azimuth angle at which the target was captured. Therefore, even when the shape of the target looks different depending on the viewing angle of the target, it is possible to correctly determine what the target is.

Embodiment 2 FIG . In the above-described first embodiment, the target is determined by comparing the entire captured silhouette of the target with the entire calculated silhouette of the target candidate. The present embodiment is characterized in that the target is discriminated with higher accuracy by further comparing the characteristic shapes of the silhouettes.

FIG. 2 shows a second target discriminating apparatus according to the present invention.
It is a block diagram showing an embodiment of. In the present embodiment, as a part of the comparison / discrimination means, the specific shape included in the cross-sectional shape calculated by the target cross-section calculating unit 14 and the target imaging extracted by the preprocessing unit 12 corresponding to the specific shape. A unique shape determination unit 21 that further compares the unique shape included in the shape data is provided.
The feature is that not only the comparison of silhouettes but also the determination result by the unique shape determination unit 21 is considered. In the present embodiment, the case where the protrusions are described as the peculiar shape showing the target structural feature is explained, but the accuracy of the discrimination is improved by comparing the number of protrusions of the silhouette in addition to the comparison of the entire silhouettes. be able to.

Next, the operation of this embodiment will be described.

First, the process until the preprocessing unit 12 extracts the target imaging shape data and the process of the target section calculation unit 14 are the same as those in the first embodiment, but in this embodiment, The digital data extracted by the processing unit 12 is sent not only to the comparison determination unit 25 but also to the unique shape determination unit 21.

The unique shape determination unit 21 calculates the number of protrusions from the target shape data sent from the preprocessing unit 12. Further, the number of protrusions is calculated from the cross-sectional shape sent from the target cross-section calculation unit 14 by the same calculation method. By calculating the number of protrusions based on the target cross-sectional shape when viewed from the position rotated based on the azimuth angle at which the target was imaged, and by using the same calculation method as that for the target shape data There is no difference in the certification as.

The comparison / determination unit 25 determines the number of protrusions of each data to be compared before comparing the imaging shape data sent from the preprocessing unit 12 with the sectional shape sent from the target section calculation unit 14. Compare. As a result of the comparison of the numbers of protrusions, if the number is not the same or the number of differences within the preset allowable range, it is determined that the target candidate to be compared is not the imaged target. Then, the sectional shape of only the target candidate that has not been determined to be not the target is compared with the imaging shape data, and the sectional shape, that is, the matching pattern target candidate whose shape can be determined to be the closest, is determined as the determination result from the determination result output line 18. Output.

As described above, according to the present embodiment, the number of protrusions representing the features of the target structure is compared, the matching pattern of the target candidates is selected, the target candidates are narrowed down, and the entire silhouette is calculated. Since I tried to compare
It is possible to accurately determine even a target or the like having a similar outer shape and difficult to determine. Further, by comparing the number of protrusions first, it is possible to shorten the overall processing time for performing the target determination.

In the present embodiment, the case where the projection is used as the unique shape showing the target structural feature has been described, but it is also possible to use other unique shapes such as depressions and the number of sides. There is no end. Further, although the silhouettes are compared after the number of protrusions is compared, the comparison may be performed in the reverse order.

Embodiment 3. In each of the above embodiments, the target is discriminated by comparing the data based on the target shape. However, in the present embodiment and thereafter, the target discrimination support information based on the information other than the shape of the target. Is generated and the target discrimination support information is taken into consideration in the comparison discrimination, thereby improving the accuracy of the target discrimination. Among them, in Embodiments 3 to 6 described below, support information generating means for generating target discrimination support information based on the support data from the above-mentioned camera or support data detecting means, shape data of target candidates and targets. Target information storage means for storing target specifying information relating to candidate feature items, and target discrimination support information together with target discrimination by comparing target imaging shape data and target candidate shape data as in the above embodiment. It is characterized in that a comparison determination means for determining a target in consideration of a comparison result with the target specifying information is provided. The present embodiment is characterized in that the infrared image is used together with the shape data to perform the comparison determination by also matching the position of the heat source that the target has, so that the target heat source position extracted based on the infrared image is targeted. The heat source position specifying unit that is obtained as the discrimination support information is provided as the support information generation unit.

FIG. 3 shows a third target discriminating apparatus according to the present invention.
It is a block diagram showing an embodiment of. The target discriminating apparatus in the present embodiment is the same as the above-mentioned image input unit
1. In addition to the preprocessing unit 12, an infrared image input unit 31, an infrared image preprocessing unit 32, and a heat source determination unit 3 as heat source position specifying means.
3 and the image alignment unit 34, the comparison determination unit 35 as the comparison determination unit, and the matching pattern storage unit 36 that stores the heat source position of the target candidate as the target specifying information as the target information storage unit. . The infrared image input unit 31 inputs, through the infrared image input line 37, an infrared image obtained by simultaneously shooting the same target from the infrared camera as the support data detecting unit when shooting the image input to the image input unit 11. The infrared image preprocessing unit 32 cuts out a target area from the input infrared image. Heat source determination unit 33
Determines the position of the heat source in the region cut by the infrared image preprocessing unit 32. The image registration unit 34 superimposes the position of the heat source determined by the heat source determination unit 33 on the target region cut out by the preprocessing unit 12. Comparison determination unit 35
Performs comparison determination with the image superimposed by the image registration unit 34.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the pre-processing unit 12 until the target area is cut out from the captured image of the target and the imaging shape data is extracted is the same as in the first embodiment. Is. The extracted target imaging shape data is sent to the image registration unit 34.

On the other hand, an infrared image photographed at the same time as the photographing of the target by the camera is inputted from the infrared image input line 37 to the infrared image input section 31. The infrared image input unit 31 converts the input infrared image into digital data and transmits it as one image to the infrared image preprocessing unit 32. The infrared image preprocessing unit 32 performs processing for cutting out a target area to be discriminated from the digital image. The cutting process is performed by the same method as the preprocessing unit 12. The cut out region is sent to the heat source determination unit 33. The heat source determination unit 33 determines the position of the heat source, which is the center of heat generation, from the spread of the luminance distribution of the target area of the cut out infrared image. The determination result is sent to the image registration unit 34.

In the image registration unit 34, the preprocessing unit 12
In the area where the target sent from the
The positions of the heat sources sent from the device are overlapped, and the generated target pattern is output to the comparison / determination unit 35 as target determination support information. Since the matching pattern storage unit 36 stores the matching pattern in which the shape data of the target candidate and the position of the heat source of each target candidate are overlapped as the original data, in the comparison / determination unit 35, from the image registration unit 34. The target pattern to be sent and the matching pattern storage unit 36
Is compared with the matching pattern of the target candidate stored in, and as a result of the comparison, the target candidate of the matching pattern that can be determined to be the closest is output from the determination result output line 18 as the determination result.

As described above, according to the present embodiment, when the comparison for the target determination is performed, the matching pattern having not only the shape but also the heat source position is used. It is possible to accurately determine the target even in a blurred captured image that is hard to stick.

Fourth Embodiment FIG. 4 is a block diagram showing a fourth embodiment of the target discriminating apparatus according to the present invention. The target discriminating apparatus according to the present embodiment includes support information for calculating a target dimension as target discriminating support information from the distance measurement data input from the distance measurement data input line 41 in addition to the image input unit 11 and the preprocessing unit 12 described above. The target dimension calculation unit 42 as a generation unit is provided. The distance measurement data is distance data from the camera to the target, which is measured at the same time as the image input to the image input unit 11 is captured. In the present embodiment, the target image capturing camera is also used as the support data detecting means. To do. Further, the matching pattern storage unit 46 in the present embodiment stores the size information of the target candidate as the target specifying information, and the comparison determination unit 45 accordingly compares the target shapes and sizes. That is, in the third embodiment, the heat source position is used as the target determination support information, but in the present embodiment, the target dimension information is used as the target determination support information.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the pre-processing unit 12 until the target area is cut out from the captured image of the target and the imaging shape data is extracted is the same as in the first embodiment. Is. The extracted target imaging shape data is sent to the comparison / determination unit 45.

On the other hand, distance measurement data from the camera to the target, which is measured at the same time as the photographing of the target, is input to the target dimension calculation section 42 from the distance measurement data input line 41. The target size calculation unit 42 calculates the size of the target region of interest from the input distance measurement data and the viewing angle of the camera. The calculated dimensions are sent to the comparison / discrimination unit 45 as target discrimination support information. Since the matching pattern storage unit 46 stores the matching pattern including the shape data of the target candidate and the size information of each target candidate as the original data, the comparison / determination unit 45 first sends it from the target size calculation unit 42. The target candidate is selected by comparing the size of the target target with the size of the target candidate stored in the matching pattern storage unit 46. Then, the shape data of the selected target candidate is compared with the target imaging shape data sent from the preprocessing unit 12, and the target result of the matching pattern whose shape is the closest can be determined as the determination result. Output from 18.

As described above, according to this embodiment, not only the shape but also the size of the target is used as the support information for the target determination when performing the comparison for the target determination. It is possible to accurately determine the target even in a blurred captured image that is difficult to determine by itself.

In the above description, by comparing the target dimensions and narrowing down the target candidates to some extent and then comparing the shapes, it is possible to shorten the overall processing time for determining the target. However, the procedure and method of the comparison / determination process performed using the target shape data and the dimension information are design matters. Further, in the process of comparing and discriminating the target dimension calculated from the imaging data and the target actual dimension, it is a design matter to allow up to what difference.

Embodiment 5. FIG. 5 is a block diagram showing the fifth embodiment of the target discriminating apparatus according to the present invention. The target discriminating apparatus according to the present embodiment is obtained via the geographic data storage unit 51 that stores geographic information and coordinate information and the target position data input line 52, in addition to the image input unit 11 and the preprocessing unit 12 described above. A geographic determination unit 53, which generates positional information of the target obtained as a result of comparison between the imaged data of the target captured by the camera and the information stored in the geographic data storage unit 51, is provided as support information generation means. Further, the matching pattern storage unit 56 in the present embodiment stores the geographic information about the target candidate as the target specifying information, and the comparison / determination unit 55 accordingly compares the shape of the target and the geographic information.
That is, the present embodiment is characterized in that the target geographical information is used as the target discrimination support information. The topographical information includes three-dimensional information such as land and sea topography in addition to map information such as flat roads. The topographical data storage unit 51 divides the entire country of Japan into meshes, for example. Topographical information and coordinate information are stored for each mesh. Further, in the present embodiment, the camera for target photographing is also used as the assistance data detecting means.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the preprocessing unit 12 to cut out the target area from the image of the target and extract the imaging shape data is the same as in the first embodiment. Is. The extracted target imaging shape data is sent to the comparison / determination unit 55.

On the other hand, the image pickup data is inputted from the target position data input line 52 to the geographical record judging section 53. The geographic determination unit 53 obtains the target position while comparing the geographic information stored in the geographic data storage unit 51 with the coordinate information based on the input imaged data. Then, the position information of the target is obtained.
The position information is, for example, vehicles if the target position is a road,
If it is on the water, it will be like a ship. The position information is sent to the comparison / determination unit 55.

The comparison / determination unit 55 compares the position information of the target target sent from the geographic determination unit 53 with the geographic information of the target candidates such as vehicles and ships stored in the matching pattern storage unit 56. Select a target candidate with. Then, the shape data of the selected target candidate and the preprocessing unit 1
2 is compared with the target imaging shape data sent from No. 2 and the target candidate of the matching pattern whose shape is judged to be the closest is output from the judgment result output line 18 as the judgment result.

As described above, according to the present embodiment, not only the shape but also the position information of the target is used as the support information for the target determination when performing the comparison for the target determination. It is possible to accurately determine the target even in a blurred captured image that is difficult to judge only by the outer shape. In the above description, the target geographic information is compared to narrow down the target candidates to some extent and then the shapes are compared, so that the overall processing time for performing the target determination can be shortened.

Sixth Embodiment FIG. 6 is a block diagram showing a sixth embodiment of the target discriminating apparatus according to the present invention. The target discriminating apparatus according to the present embodiment includes, in addition to the image input unit 11 and the preprocessing unit 12 described above, an electronic device specification data storage unit 61 that stores specification data of an electronic device that can be mounted on the target, and a target. The target mounted electronic device specified on the basis of the radio wave data sent from the support data detecting means for receiving the radio wave radiated by the user via the radio wave data input line 62 and the specification data stored in the electronic device specification data storage unit 61. An onboard electronic device determination unit 63 that generates a device as target determination support information is provided as support information generation means. Further, the matching pattern storage unit 66 in the present embodiment is
The electronic device mounted on each target candidate is stored as target specifying information, and the comparison / determination unit 65 accordingly compares the shape of the target and the type of electronic device. That is, the present embodiment is characterized in that the target on-board electronic device is used as the target discrimination support information.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the preprocessing unit 12 until the target area is cut out from the captured image of the target and the imaging shape data is extracted is the same as in the first embodiment. Is. The extracted target imaging shape data is sent to the comparison / determination unit 65.

On the other hand, the measured radio wave data of the photographing target is input from the radio wave data input line 62 to the mounted electronic device determination section 63. Based on the input radio wave data, the on-board electronic device determination unit 63 identifies the on-board electronic device considered to be the target target radiated radio wave from the electronic device specification data stored in the electronic device specification data storage unit 61. To do. The identified on-board electronic device information is used as the target discrimination support information by the comparison / discrimination unit 65.
Sent to

The comparison / determination unit 65 compares the target on-vehicle electronic device information sent from the on-vehicle electronic device determination unit 63 with the target candidate matching pattern stored in the matching pattern storage unit 66, that is, the electronic device information. The target candidates are narrowed down to some extent by. Then, the narrowed down target candidate shape data is compared with the target imaging shape data sent from the preprocessing unit 12, and the target candidate of the matching pattern whose shape is the closest can be used as the determination result as a determination result output line. Output from 18.

As described above, according to the present embodiment, when performing comparison for target determination, not only the shape but also the mounted electronic device specified based on the target radiated radio wave can be used for target determination. Since the information is used as the support information, it is possible to accurately determine the target even in a blurred captured image that is difficult to determine only by the outer shape. Further, in the above description, by comparing the target electronic device information, narrowing down the target candidates to some extent, and then comparing the shapes, it is possible to shorten the overall processing time for performing the target determination.

Embodiment 7. FIG. 7 is a block diagram showing a seventh embodiment of the target discriminating apparatus according to the present invention. The target discriminating apparatus according to the present embodiment includes, in addition to the image input unit 11 and the preprocessing unit 12 described above, a speed performance specification data storage unit as a speed performance specification data storage unit that stores speed performance information of each target candidate. 71, target speed performance selection means for selecting target candidates that can be determined by comparing speed data input via the target speed data input line 72 with speed performance information stored in the speed performance specification data storage unit 71. Target speed / performance sorting unit 73, a matching pattern storage unit 76 as a target shape data storage unit that stores target candidate shape data, and a target speed / performance sorting unit of the shape data stored in the matching pattern storage unit 76. Shape data for target candidates selected by 73,
The pre-processing unit 12 includes a comparison / determination unit 75 as a comparison / determination unit that determines the target by comparing it with the target imaging shape data extracted. What is characteristic of this embodiment is that the target is determined in consideration of not only the shape of the target but also the moving speed of the target.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the preprocessing unit 12 until the target area of interest is cut out from the image of the target and the imaging shape data is extracted is the same as in the first embodiment. Is. The extracted target imaging shape data is sent to the comparison / determination unit 75.

On the other hand, the speed data of the same target measured simultaneously with the shooting of the target by the camera is used as the target speed data input line 7
Input from 2 to the target speed performance selection unit 73. Based on the input target speed data, the target speed / performance sorting unit 73 considers the target target moving speed from the speed / performance specification data of the target stored in the speed / performance specification data storage unit 71. Select target candidates that can be used. For example, when the speed data is 100 km / h, target candidates whose maximum speed is less than 100 km / h are not selected. In this way, target candidates are narrowed down based on the actual moving speed of the target. The selected target candidates are sent to the comparison / determination unit 75.

The comparison / determination unit 75 selects and extracts the matching pattern, that is, the shape data of the target candidate stored in the matching pattern storage unit 76 based on the target candidate sent from the target speed / performance selection unit 73, and the preprocessing unit. 1
The target candidate of the matching pattern that can be judged to have the closest shape is output from the judgment result output line 18 as a judgment result.

As described above, according to the present embodiment, when the comparison for the target determination is performed, the target candidates are narrowed down in advance based on the actual moving speed of the target, and only the narrowed down target candidates are selected. Since the comparison and determination processing is performed based on the shape, even when there are targets having similar outer shapes, the targets can be accurately determined when there is a difference in moving speed. Further, by narrowing down the target candidates to some extent and then comparing the shapes, it is possible to reduce the overall processing time for performing the target determination.

Embodiment 8. FIG. 8 is a block diagram showing the eighth embodiment of the target discriminating apparatus according to the present invention. In addition to the image input unit 11 and the pre-processing unit 12 described above, the target discriminating apparatus according to the present embodiment has a target placement as a target placement detecting unit that obtains the placement status of a plurality of targets lined up in a captured image. A calculation unit 81 and a formation / arrangement data storage unit 82 as a target arrangement data storage unit that stores formation formation information such as formations and formations that target candidates can take.
And the formation / arrangement determination unit 83 as a target arrangement determination unit that identifies formation formation information corresponding to the arrangement situation obtained by the target arrangement calculation unit 81 from the formation arrangement information stored in the formation / arrangement data storage unit 82. And have. Further, the matching pattern storage unit 86 in the present embodiment is a target information storage unit that stores shape data of target candidates and formation formation information regarding formations, formations and the like taken by the target candidates, and the comparison determination unit 85 is described above. The comparison / determination means determines the target based on not only the target shape data but also the formation formation information specified by the formation / placement determination unit 83 and the formation formation information stored in the matching pattern storage unit 86.

In the present embodiment, when the target forms a formation or the like, the target determination is performed by using the arrangement information of the formation of the target as the target determination support information.

Next, the operation of this embodiment will be described.

First, the processing performed by the image input unit 11 and the preprocessing unit 12 until the target area is cut out from the image of the target and the imaging shape data is extracted is the same as in the first embodiment. Is. However, in the present embodiment, the extraction process is performed for a plurality of targets. The extracted target imaging shape data is sent to the comparison / determination unit 85 and the target placement calculation unit 81.

The target placement calculation unit 81 calculates the placement relationship of the regions of each target target sent from the preprocessing unit 12, and sends it to the formation / placement determination unit 83. In the formation / placement determination unit 83,
Based on the input placement relationship of the targets, formation formations that are considered for the placement of the target target are determined from formation formation information such as formations and formations of possible target candidates stored in the formation / placement data storage unit 82. In this determination process, it is not always necessary to specify a single formation. The formation formation information determined and selected is sent to the comparison / determination unit 85.

In the comparison / determination unit 85, the formation / placement determination unit 8
From the formation formation information of the target target sent from No. 3, it is compared with the matching pattern of the target candidate stored in the matching pattern storage unit 86, that is, formation formation information. Then, as a result of the comparison, for the selected target candidate, the shape data of the target candidate stored in the matching pattern storage unit 86 and the imaging shape data of the target target sent from the preprocessing unit 12 are compared, The target candidate of the matching pattern that can be determined to have the closest shape is output from the determination result output line 18 as the determination result.

As described above, according to the present embodiment, when the formations are formed by a plurality of targets when performing the comparison for the target discrimination, the target candidates are formed based on the formation formation information of the targets. Since the target is narrowed down in advance and the comparison and determination processing based on the shape is performed only on the narrowed down target candidates, the target determination can be accurately performed even in a blurred captured image that is difficult to determine only by the outer shape. Further, by narrowing down the target candidates to some extent and then comparing the shapes, it is possible to reduce the overall processing time for performing the target determination. Further, when the formation has a characteristic, even when there are targets having similar outer shapes, it is possible to accurately discriminate between the targets.

In the above example, all the targets forming the formation are the same. However, when forming different targets, the target is formed by giving information such as the type of the target forming the formation. It can be used as discrimination support information.

Ninth Embodiment In each of the above-described embodiments, not only comparison of general target shape data but also accurate determination of a target based on a silhouette of a shape, a structural feature, or target determination support information used separately has been described. In the present embodiment, as shown in FIG. 9 and FIG. 10, a target discriminating apparatus for discriminating a target using all the target discriminating support information such as the above-mentioned target heat source position and dimensions is provided. Therefore, the matching pattern storage unit 96 in the present embodiment stores each target specifying information and the like described in each of the above-described embodiments in addition to the target candidate shape data, and the comparison determination unit 95 determines the target shape data and the target shape data. The target is discriminated based on the result of comparison between the target specifying information and the like stored in the matching pattern storage unit 96 and the target discriminating support information and the like described in the above embodiments.

Since the operation of this embodiment is composed of a combination of those described in each of the above embodiments, the description of the operation of each component will be omitted. Comparison determination unit 9
Which target discrimination support information or the like is used in order to narrow down the target candidates in the comparison and discrimination processing according to 5 is merely a design matter. As a result, in this embodiment, the same effect as that of each of the above embodiments can be obtained.

In this embodiment, all the constituent elements described in each of the above-mentioned embodiments are provided, but it is possible to appropriately select and combine them according to the field to which the apparatus is applied and the application. Good.

[0076]

According to the present invention, the imaged shape data of the target actually imaged is compared with the sectional shape of the target when viewed from the position rotated based on the azimuth angle of the imaged target, that is, the silhouette. Therefore, even when the shape of the target looks different depending on the viewing angle of the target, it is possible to correctly determine what the target is.

Further, not only the comparison of the entire silhouettes,
Since the comparison of the peculiar shapes is also performed, the target can be discriminated with higher accuracy. In addition, if the comparison of the entire silhouette is performed after the target candidates are narrowed down by performing the comparison based on the unique shape such as the number of protrusions, it is possible to shorten the entire processing time for performing the target determination.

Further, since the matching pattern having not only the shape but also the position of the heat source is used when the comparison for the target discrimination is performed, the target can be accurately targeted even in the blurred image which is difficult to judge only by the outer shape. It is possible to make a distinction. Further, even when there are targets having similar outer shapes, the heat source positions are separately used as the target determination support information for comparison, so that it is possible to accurately determine the targets.

Further, when the comparison for the target discrimination is performed, not only the shape but also the target dimension is used as the support information for the target discrimination. Can also accurately determine the target. Further, even when there are targets having similar outer shapes, the target dimensions are separately used as the target discrimination support information for comparison, and therefore it is possible to accurately discriminate between the targets.

Further, when the comparison for the target discrimination is performed, not only the shape but also the position information of the target is used as the support information for the target discrimination. The target can be determined with high accuracy. Further, even when there are targets having similar outlines, the target geographic information is separately used as the target determination support information for comparison, so that it is possible to accurately determine the targets.

Further, when the comparison for the target discrimination is performed, not only the shape but also the mounted electronic device specified based on the target radiated radio wave is used as the support information for the target discrimination. With this, it is possible to accurately determine the target even in a blurred captured image that is difficult to determine. Further, even when there are targets having similar outlines, the target geographic information is separately used as the target determination support information for comparison, so that it is possible to accurately determine the targets even when the targets are distinguished from each other. Become.

Further, when performing the comparison for the target discrimination, the target candidates are narrowed down in advance based on the actual moving speed of the target, and only the narrowed down target candidates are subjected to the comparison discrimination processing based on the shape. Even when there is a target having a similar outer shape, the target can be accurately determined when the moving speeds are different. Further, by narrowing down the target candidates to some extent and then comparing the shapes, it is possible to reduce the overall processing time for performing the target determination.

In addition, when making a comparison for the purpose discrimination, if the formation is composed of a plurality of targets, not only the shape but also the formation formation information of the target should be used as the support information for the purpose discrimination. Therefore, it is possible to accurately determine the target even in a blurred captured image that is difficult to determine only by the outer shape. Further, when the formation has a characteristic, even when there are targets having similar outer shapes, it is possible to accurately discriminate between the targets. Further, if the target candidates are narrowed down to some extent by the formation formation information and then the shapes are compared, the overall processing time for the target determination can be shortened.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing a first embodiment of a target discriminating apparatus according to the present invention.

FIG. 2 is a block configuration diagram showing a second embodiment of a target discriminating apparatus according to the present invention.

FIG. 3 is a block configuration diagram showing a third embodiment of a target discriminating apparatus according to the present invention.

FIG. 4 is a block configuration diagram showing a fourth embodiment of a target discriminating apparatus according to the present invention.

FIG. 5 is a block configuration diagram showing a fifth embodiment of a target discriminating apparatus according to the present invention.

FIG. 6 is a block configuration diagram showing a sixth embodiment of a target discriminating apparatus according to the present invention.

FIG. 7 is a block configuration diagram showing a seventh embodiment of a target discriminating apparatus according to the present invention.

FIG. 8 is a block diagram showing an eighth embodiment of the target discriminating apparatus according to the present invention.

FIG. 9 is a block configuration diagram showing a part of a ninth embodiment of a target discriminating apparatus according to the present invention.

FIG. 10 is a block diagram showing another part of the ninth embodiment of the target discriminating apparatus according to the present invention.

FIG. 11 is a configuration diagram of a conventional character recognition device for image recognition.

[Explanation of symbols]

11 image input unit, 12 pre-processing unit, 13 three-dimensional data storage unit, 14 target section calculation unit, 15, 25, 35,
45, 55, 65, 75, 85, 95 Comparison determination unit, 1
6 image input line, 17 shooting condition data input line, 18
Discrimination result output line, 21 Unique shape determination unit, 31 Infrared image input unit, 32 Infrared image preprocessing unit, 33 Heat source determination unit, 34 Image alignment unit, 36, 46, 56, 6
6,76,86,96 matching pattern storage unit, 3
7 Infrared image input line, 41 Distance measurement data input line, 42
Target dimension calculation unit, 51 Topographic data storage unit, 52 Target position data input line, 53 Topographical determination unit, 61 Electronic equipment specification data storage unit, 62 Radio wave data input line, 63 Onboard electronic equipment determination unit, 71 Speed performance specifications Data storage,
72 target speed data input line, 73 target speed performance selection unit, 81 target placement calculation unit, 82 formation / placement data storage unit, 83 formation / placement determination unit.

Claims (10)

[Claims] 1. A target discriminating device for discriminating a target displayed in an image photographed by an image pickup means, a target shape data extracting means for obtaining imaged shape data of a target imaged by said image pickup means, and a three-dimensional target candidate. A target shape data storage means for storing data, an azimuth in which the target is located is calculated from a positional relationship between the image pickup means and the target at the time of photographing, and a sectional shape when the target is viewed from the azimuth angle is stored in the target shape data storage. Comparing the target cross-section calculation means for calculating based on the three-dimensional data stored in the means, the target imaging shape data extracted by the target shape data extraction means, and the cross-sectional shape calculated by the target cross-section calculation means. And a target discriminating device for discriminating a target according to the above. 2. The comparison / discrimination means is included in the specific shape included in the cross-sectional shape calculated by the target cross-section calculating means and the target imaging shape data extracted by the target data extracting means corresponding to the specific shape. The target discriminating apparatus according to claim 1, further comprising a comparison with the unique shape. 3. The target discriminating apparatus according to claim 2, wherein the peculiar shape is a protrusion. 4. A target discriminating apparatus for discriminating a target displayed in an image photographed by an image pickup means, the assistance data detecting means for detecting assistance data for the same target at the same time as the image pickup means, and the target imaged by the image pickup means. Target shape data extraction means for obtaining the imaged shape data of the target, support information generation means for generating the target discrimination support information based on the support data by the support data detection means, and the target regarding the shape data of the target candidate and the characteristic items of the target candidate. Target information storage means for storing specific information, and a target is discriminated based on a comparison result between the target imaging shape data and the target candidate shape data and a comparison result between the target discrimination support information and the target specific information. And a comparing and determining means, for determining the target based on the target imaging shape data and the target support information. And a target discriminating device. 5. The support information generating means includes heat source position specifying means for obtaining target heat source position extracted based on an infrared image of the target by the support data detecting means as target discrimination support information, The target information storage means stores the heat source position of the target candidate as target specifying information, and the comparison and determination means, the target heat source position generated by the support information generation means, and the target candidate stored by the target information storage means. The target discriminating apparatus according to claim 4, wherein the target position is compared with a heat source position. 6. The target size calculation unit, wherein the support information generation unit calculates a target size as target determination support information from distance measurement data up to the target obtained based on imaging data obtained by copying the target by the support data detection unit. The target information storage means stores dimension information of target candidates as target specifying information, and the comparison and determination means stores the target dimensions calculated by the support information generation means and the target information storage means. The target discriminating device according to claim 4, wherein the target discriminating device compares the dimensional information of the target candidate to be processed. 7. The support information generating means includes a geographic data storage section for storing geographic information and coordinate information, imaging data showing a target by the support data detecting means, and information stored in the geographic data storage section. The topographical information determination unit that generates the position information of the target obtained as a result of the target determination support information, and the target information storage unit stores the topographical information about the target candidate as the target identification information. The target discriminating device according to claim 4, wherein the means compares the position information of the target calculated by the support information generating means with the geographic information of the target candidate stored in the target information storing means. 8. The electronic device specification data storage unit for storing the specification data of an electronic device that can be mounted on the target, and the radio wave data radiated by the target detected by the support data detection unit. And a mounted electronic device determination unit that generates target mounted electronic device specified as target discrimination support information based on the specification data stored in the electronic device specification data storage unit, and the target information storage unit includes An electronic device mounted on the target candidate is stored as target specifying information, and the comparison / determination unit stores the target electronic device calculated by the support information generating unit, and a target candidate electronic device stored by the target information storage unit. 5. The target discriminating device according to claim 4, characterized in that 9. A target discriminating apparatus for discriminating a target displayed in an image photographed by an image pickup means, a target shape data extracting means for obtaining image pickup shape data of the target imaged by said image pickup means, and a speed performance of each target candidate. Speed performance specification data storage means for storing information, speed data for the same target detected at the same time as the imaging means, and target candidates that can be determined by comparison with speed performance information stored by the speed performance specification data storage means Target speed performance selection means for selecting the target speed performance selection means, target shape data storage means for storing the shape data of the target candidates, and target candidates selected by the target speed performance selection means from the shape data stored in the target shape data storage means. By comparing the shape data for the target image pickup shape data extracted by the target shape data extraction means. Target discriminating apparatus characterized by having a comparison discriminating means for discriminating. 10. A target discriminating device for discriminating a target displayed in an image photographed by an image pickup means, and target information storage means for storing shape data of the target candidate and formation formation information regarding the formation, formation etc. taken by the target candidate. Target formation data storage means for storing formation formation information such as formations and formations that the target candidate can take; target shape data extraction means for obtaining each image pickup shape data of a plurality of targets imaged by the image pickup means; Target placement detection means for obtaining the placement situation of a plurality of targets imaged by means, and formation formation information corresponding to the placement situation obtained by the target placement detection means of formation formation information stored in the target placement data storage means. A target placement determining unit that identifies the target placement shape, a comparison result of the target imaging shape data and the target candidate shape data, and the target placement determination A target discriminating device comprising: comparison discriminating means for discriminating a target based on a comparison result of the formation formation information specified by the setting means and the formation formation information stored in the target information storage means.