JP2854465B2 - Moving object tracking method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The disclosed technology is aimed at moving objects such as a tank with a camouflage such as a military target or a case where it loses its direction or is difficult to distinguish from a background of a stray car such as a missing vehicle. It belongs to the technical field of a system in which a video camera (television camera) extracts a target on an image and locks on for the next predetermined operation.

[0002]

2. Description of the Related Art As is well known, when a missile is fired at a hostile target such as a camouflage tank, the distress or direction of the missile is lost, or when a rescue supply or the like is sent to a lost or stray vehicle or the like, the moving object is used. It is necessary to take certain actions with certainty as a target, and with the recent improvement in electronic technology, coping techniques with microelectronic systems have been developed.For example, automatic tracking for athletes and balls such as sports broadcasts, storefronts, etc. It is also used in the consumer and defense industries, such as vending machines, or ticket sales and ticket sales at train stations and event venues. For example, Japanese Unexamined Patent Publication No. 61-284681. Invention and Japanese Patent Application Laid-Open No. 63-138283,
Japanese Patent Publication No. 140598 has been developed.

[0003] In the search for enemies in the air and the enemies on the sea, the image of a moving object captured by a video camera (television camera) is processed as a target brightness contrast beyond the visual recognition area with respect to the physiological limit by the naked eye. Has been put to practical use.
No. 38283 discloses the invention.

[0004] By the way, a moving object tracking method based on luminance contrast processing on such an image has a very low contrast in the target, such as a monotone background such as in the air or on the sea. The numbered ticket reading system and the like have a clear reading method, so that the image is converted to a black-and-white (monochrome) image, and a certain effect can be obtained as a measure for processing data.

[0005]

However, the contrast of the black and white (monochrome) image is as follows.
For example, when the luminance distribution is binarized by a certain threshold value and data processing is performed, for example, in the case of a ground moving object such as a tank with a camouflage, in a background grassland, desert, forest, etc. Since the tracking of the ground moving object is performed, the influence of the shade of sunlight is remarkably large, and there is a disadvantage that the data is significantly influenced by a change in luminance.

Therefore, the use of such a mobile object tracking system is limited to a search for a target in the air or the sea in the background with little change in luminance such as a search for an enemy in the air or the sea.
It is difficult to become a static mobile tracking system that is not affected by weather and terrain conditions, and it lacks mobility as a military action or search response means.

That is, for example, as shown in FIG. 5, when a tank 3 having a camouflage as a hostile moving body on a real ground 1 is moving on a grassland such as a tree 4 as a background. A missile launcher 6 intercepting through a barrier 5 such as a bank and a video camera (television camera) 8 on the barrier 5 such as a bank searching for capturing the tank 3 as a target are set at the rear of the missile launcher 6. In a mode in which the tank 3 is electrically connected to the control vehicle 7 and faces the control vehicle 7, the video camera (television camera) 8 picks up an image of the tank 3
As shown in FIG. 6, the target 3 'is recorded together with the background 4' of the tree on the image plane 9 in the control panel of the control car 7 as shown in FIG. At this time, as shown in FIG. 7, if the luminance of the vertical section in FIG. 6 of the three primary colors (red R, green G, blue B) is taken on the vertical axis and the image width is taken on the horizontal axis, red (R), green The luminance distributions of (G) and blue (B) are as shown in graphs 10, 11, and 12.

Note that 3 ″ represents a luminance distribution of the target 3 ′.

If the hue of the target 3 'is a color that is more conspicuous than that of the background 4', the brightness of the component is increased. Although it is possible to capture 3 ′, especially when the ground moving body is a camouflage tank 3 or the like, it is located between multicolored backgrounds such as grasslands, trees, and forests 4 as a background. However, in the case where the camouflage is applied in the operation, there is a difficulty that the target is difficult to identify on the opposing attacking side (it is originally a camouflage for that purpose).

In extracting the target 3 ', the luminance difference is unstable and not remarkable in a sunshine portion or a shade portion. Therefore, the extraction of the target 3' through the binarized data from the image is difficult. There is a demerit that remarkable difficulty accompanies the noise removal at the time of electronic processing and the electronic processing control becomes extremely complicated.

[0011]

SUMMARY OF THE INVENTION An object of the invention of this application is to focus on a target extraction process out of two target extraction processes and tracking calculation processes which are roughly considered as algorithms of a mobile object tracking system.
The tracking of a ground moving object used for military, search, rescue, etc. based on the prior art described above is performed when a target is determined by a hue extraction process based on luminance contrast of a color image via a video camera (television camera). A technical task to solve the problem of saliency and instability of binarized data, and search in military industry, emergency industry, etc. by increasing the accuracy of target hue extraction of captured images by luminance ratio An object of the present invention is to provide an excellent ground moving object tracking method that benefits the technical application field.

[0012]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the construction of the invention of the present application for capturing the images of the moving body and its background through a camera is described. In the moving object tracking method for electronically processing the image and determining the moving object as a target,
Using a color image as the image, the electronic processing of the image is divided into primary color (red (R), green (G), and blue (B)) images, and the luminance of each primary color image is calculated as the total image luminance of the color image. The luminance ratio image is obtained for each of the primary colors by dividing by the above, and a binarization process is performed for each of the primary colors using the luminance data of the luminance ratio image. Extracting a region recognized as a target as the moving object is a base invention. Thus, an image of the moving object and its background is taken through a camera, the images are electronically processed, and the moving object is set as the target. In the moving object tracking method which is determined as follows, a color image is used as the image, and the image is divided into primary color (red (R), green (G), and blue (B)) images. For each of the primary colors, 2 is used for each of the primary colors using the luminance difference image luminance data obtained by calculating the luminance difference. A second invention is to perform a binarization process and extract, as the moving body, a region recognized as a target in all of the binary images for each of the primary colors. , And technical measures were taken to obtain the average and deviation of the density.

[0013]

A camouflage tank that moves in a multicolored background such as a grassland or a forest in a military operation, or loses direction due to distress, or is fixed to the ground of a moving vehicle of a stray car, or The moving object is imaged together with the background by a video camera (television camera) or the like provided on a seeker of the missile warhead, a gate is applied to a target extraction area of the image captured by the central processing unit, and the original image is red (R). , Green (G), and blue (B) images, and a luminance ratio image for each of the primary color red (R), green (G), and blue (B) images is calculated by a computer. Then, the average and the variance of the deviation of the target hue are obtained, binarized based on the deviation, the theoretical product of the binarized data of each primary color (red R, green G, blue B) is taken, and isolated points are removed. Also, by performing the contraction / expansion operation, the target is raised and determined, What is to take the tracking point of the target by the area center of gravity or the like, in which certainly is tracking mobile in the background on the ground was certainly possible as.

[0012]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of the present invention;

In the illustrated embodiment, a target 23 of a color different from the background is placed in a background color 22 on the image plane shown in FIG. 4, and a shadow portion 17 is added to see the influence of the shade of sunlight. This is an aspect, and the image is captured by a video camera (television camera) of an image recording unit in the target extraction processing system in FIG.

Then, as shown in FIG. 4, an original image is captured through a filter, and the image is divided into red (R), green (G), and blue (B) images to extract the hue of the target 23. Then, the position of the target 23 becomes unclear). Therefore, the gate 16 is applied to the target 23 in advance, and the red (R), green (G), and blue (B) luminance ratio images 18, 19,
20.

Here, red (R), green (G),
The blue (B) luminance ratio images 18, 19, and 20 are separated even if the actual moving body 3 is a red color (R) similar to the background due to camouflage or the like. This is because the characteristics of the hue of the target 23 are intended to be captured by using green (G) and blue (B) together since the ratio of G and B) is slightly different from the background.

Then, red (R), green (G), and blue (B) images 18 ', 19 having a luminance ratio for each primary color (R, G, B) are divided by the full screen luminance and each image division. The average and the deviation of the density within the range of the gate 16 (target 23) are obtained for ', 20' so that the target 23 can be extracted.

IX in the original image 22 of FIG.
−IX At the same position as the luminance section, each luminance ratio image 18 ′, 1
The variance value of the deviation that can be binarized so that the target can be extracted at 9 ′ and 20 ′ is obtained, and then the luminance ratio images 18 ′, 19 ′ and 20 ′ are binarized by the threshold value obtained therefrom (image 1
8 ″, 19 ″, 20 ″). Then, a logical product is obtained from each of the binarized images (18 ″, 19 ″, 20 ″), and operations such as removal of isolated points are performed. Extract the goal 23 '.

Depending on the design, it is also possible to determine the area, the center of gravity, etc., and provide a tracking point for the target 3.

Regarding the luminance distribution and the luminance ratio in the process, as shown in FIG. 8, for example, in a red (R) image, the luminance distribution depends on the shadow of light, and the target 23 '' is separated from the background. Trying to set the threshold to 150,
Only a part of the target 23 ″ can be extracted, and if this is processed into a luminance ratio image as shown in FIG. 4, a threshold 150 is obtained as shown in FIG. 3 regardless of the shading of light. Goal 2 clearly
3 '''can be extracted.

In the invention of this application, the luminance of each of the red (R), green (G), and blue (B) images of the color image is used.

That is, utilizing the color factor, the influence of sunlight shading is small, it is easy to capture the luminance characteristics of the target 23 ', and it is possible to cope with a more complex and multicolored background. This makes it possible to extract the target more accurately.

Thus, the target moving body 3 and the background 4
As long as and are not exactly the same color, hue can be extracted irrespective of the shading due to sunlight.

According to an experiment, the target extraction processing mode using a (monochrome) image utilizing the luminance difference of the conventional mode and the target 23 extraction processing mode using a multi-color image based on the luminance ratio of the present invention are described in the present application. It was confirmed that in the case of the invention, the decision by the clear lifting of the target 23 'as the target in the logical product was made more reliably.

It should be noted that the embodiments of the invention of this application are not limited to the above-described embodiments. For example, in this system, red (R), green (G), and blue (B) are obtained at all image luminances. Since the number of colors to be used is three times as large as that of an image and one color of monochrome, various modifications are possible such as performing subtraction instead of division and extracting the target 23 'using the luminance difference. Embodiments can be employed.

Also, the present invention can be applied not only to a ground moving body in a military operation but also to a search for a lost or lost vehicle in a mountainous area, a search for a lost or lost vehicle, and a supply of emergency supplies. That is.

[0026]

As described above, according to the invention of this application, basically, in a military operation or the like, an attack operation or the like on a moving object such as a camouflage tank moving on a background of grassland or forest on the ground in a military operation or the like. To perform target extraction from a captured image of a body via a video camera (television camera), the captured image is divided into three primary colors, red (R), green (G), and blue (B),
By performing the binarization process and calculating the luminance ratio of each of the divided images when extracting the target from the data, the luminance ratio is stably remarkable without being affected by the luminance of a sunshine portion or a shade portion due to sunlight. An excellent effect is achieved in that a desired target can be determined and a predetermined operation for the target can be performed reliably.

The system used in the invention of this application can also employ a video camera (television camera) (for example, a missile seeker or the like) fixed on the ground or mounted on a moving body. It has the advantage of a wide range of applications that can be used not only for operations but also for searching for distressed vehicles in mountainous areas.

Also, there is an effect that the advantage that the search for a moving object that cannot be visually recognized by the naked eye can be performed electronically via a video camera (television camera) can be fully utilized.

[Brief description of the drawings]

FIG. 1 is a processing flowchart of an image according to the invention of this application.

FIG. 2 is a schematic diagram of a system used.

FIG. 3 is a graph showing a luminance ratio of red (R).

FIG. 4 is a flowchart of image processing of the divided three primary colors (R, G, B).

FIG. 5 is a schematic cross-sectional view of a moving object and a video camera (television camera) facing each other.

FIG. 6 is a schematic front view of all images.

FIG. 7 is a schematic luminance sectional view of an original image.

FIG. 8 is a graph of a luminance ratio of red (R) based on the prior art.

[Explanation of symbols]

 8 Camera 3 Moving object 4 Background 9 Color image

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-4-184283 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) F41G 7/22 G01S 3/78

Claims (1)

(57) [Claims] Imaging the background and the moving body by 1. A camera,
The resulting image electronically processed, in a mobile tracking method so as to determine the mobile as a target, using a color image on the image, the primary color the electronic processing of the color image (red (R), green ( G), blue (B)) image, and a luminance ratio image is obtained for each of the primary colors by dividing the luminance of each primary color image by the total image luminance of the color image. A binarization process for each of the primary colors, and extracting a region recognized as a target in all of the binarized images for each of the primary colors as the mobile object.