JP2007072665A - Object discrimination device, object discrimination method and object discrimination program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently discriminate an object by a pattern matching method. <P>SOLUTION: This object discrimination device stores a plurality of templates for each distance to an object to be discriminated. Then, when detecting the object to be discriminated from an image by a background difference method or a conspicuous calculation method, the object discrimination device detects the distance to the object, and selects the template corresponding to the distance. Then, the object discrimination device discriminates the object by applying the template to the object detected from the image as the object to be discriminated. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an object discriminating apparatus, an object discriminating method, and an object discriminating program for applying a template to a predetermined area in an image and discriminating an object in the area.

  2. Description of the Related Art Conventionally, for example, a technique for detecting an object that can be an obstacle to vehicle travel and avoiding a collision with the object has been developed. Therefore, in order to detect such an object that is an obstacle around the vehicle, there is a technique for photographing the periphery of the vehicle using a camera or the like and processing the captured image. As a method for detecting an object projected in the image, for example, there is a background subtraction method. In such a background difference method, each pixel of the current image currently input and the immediately preceding image input immediately before is compared, and an object that is not seen in the immediately preceding image is found in the region where the difference is extracted. It is assumed that the object exists, and the object is detected from the image as a candidate for a travel obstacle. If such a method is used, an object that is a candidate for a running obstacle is detected, but it is still unclear what the detected object is.

  Thus, there is a pattern matching method (template matching method) as an image processing method for determining what the object is as a candidate for a running obstacle detected from within an image (see, for example, Patent Document 1). In this pattern matching method, a template is applied to an object that is a candidate for a running obstacle detected from within an image, and whether or not the object is a specific object (for example, a pedestrian) matches the template and the candidate object. Image processing is performed to determine the degree.

JP 59-226981 A JP 2002-99997 A

  By the way, the above-described conventional technique has a problem in the efficiency of object discrimination by the pattern matching method as described below.

  That is, in the conventional pattern matching method described above, discrimination is performed by applying all templates without selecting a template according to the distance to the object to be discriminated from templates stored in advance. There is a problem with its efficiency.

  In addition, when discriminating an object projected in an image captured by an in-vehicle surveillance camera or the like by a pattern matching method, it is not possible to efficiently discriminate the object from images input at a fixed interval. There may be a case where the object discrimination in the image is not in time, and there is a problem in predicting the danger while driving the vehicle and avoiding the danger as soon as possible.

  There is also a technique for recognizing a pedestrian on a road by template matching (see, for example, Patent Document 2). According to this, a technique is disclosed in which a road recognized in an image is divided into predetermined regions and a template suitable for each region is applied.

  However, in such a method, since it is premised on detecting a road area, for example, in a parking lot or at night, it is difficult to detect a road area in an image. There is a problem that template matching according to the road area cannot be performed. Furthermore, since template matching is performed even if no object exists in the image, there is a problem that processing becomes inefficient.

  Accordingly, the present invention has been made to solve the above-described problems of the prior art, and provides an object discrimination device, an object discrimination method, and an object discrimination program capable of efficiently performing object discrimination by a pattern matching technique. The purpose is to provide.

  In order to solve the above-described problems and achieve the object, the invention according to claim 1 is an object discrimination device that applies a template in an image and discriminates a detected object, up to the object to be discriminated. A template storage means for storing a plurality of templates for each distance; a template selection means for detecting a distance to the object from the templates stored in the template storage means; and selecting a template corresponding to the distance; And an object discriminating unit that discriminates the object by applying the template selected by the template selecting unit in the image.

  In the invention according to claim 2, in the above invention, the template storage unit stores a plurality of templates for each distance to the object to be determined and the type of the object, and the template selection unit includes: A distance to an object in the image is detected from the templates stored in the template storage means, a template corresponding to the distance is selected for each type of the object, and the object determination means is the template selection Various types of templates selected by the means are applied in the image to discriminate the object.

  According to a third aspect of the present invention, in the above invention, the template storage means stores a plurality of templates for each distance to the object to be determined and the size of the object, and the template selection means Detecting the distance and size to the object from the templates stored in the template storage means, selecting a template corresponding to the distance and size, and the object determination means by the template selection means The selected template is applied to the image to discriminate the object.

  According to a fourth aspect of the present invention, in the above invention, the template storage unit stores a group of templates having different aspects with respect to the object to be determined for each predetermined distance, and the template selection unit includes: The distance to the object is detected from the templates stored in the template storage means, and a group of templates having different modes corresponding to the distance is selected, and the object determination means is selected by the template selection means. The group of different templates having different modes is applied to the image to discriminate the object.

  According to a fifth aspect of the present invention, in the above invention, when there are a plurality of objects to be discriminated in the image, from which object the template is preferentially applied. And priority control means for controlling the templates selected by the template selection means so as to be preferentially applied within the image based on the priority information. .

  Further, in the invention according to claim 6, in the above invention, the priority information prescribes that the template is sequentially applied preferentially from the object having the smallest distance to the object to be determined, The priority control means controls the templates selected by the template selection means so as to be preferentially assigned based on the priority information from the object having the smallest distance to the object to be determined. It is characterized by that.

  The invention according to claim 7 is the above invention, wherein the priority control means selects the template selected by the template selection means based on the priority information until the next image is input. The control is performed so as to be applied to a predetermined area.

  The invention according to claim 8 is the object discrimination method for discriminating a detected object by applying a template in an image according to the above invention, wherein a plurality of templates are provided for each distance to the object to be discriminated. A template storing step for storing the template, a template selecting step for detecting a distance to the object from the templates stored in the template storing means, and selecting a template corresponding to the distance, and selecting by the template selecting means And an object discriminating step for discriminating the object by applying the template to the image.

  According to a ninth aspect of the present invention, in the above-described invention, in the object determination program for applying a template in an image and causing a computer to execute a method of determining a detected object, the object to the object to be determined is determined. A template storage procedure for storing a plurality of templates for each distance; a template selection procedure for detecting a distance to the object from the templates stored in the template storage means; and selecting a template corresponding to the distance; An object discriminating procedure for discriminating the object by applying the template selected by the template selecting means in the image is executed by a computer.

  According to the first, eighth, and ninth aspects of the invention, the object discriminating apparatus stores a plurality of templates for each distance to the object in the image to be discriminated, and calculates the distance to the object displayed in the image. Since the object is discriminated by applying the template after detecting and selecting a template according to the distance, the object discrimination by the pattern matching method can be performed efficiently. Furthermore, since this object discrimination device detects an object in the image and performs pattern matching based on the distance to the detected object, template matching can be performed efficiently even if the road region cannot be specified. Regardless of whether or not is detected, the processing efficiency can be improved as compared with the template matching.

  Further, according to the invention of claim 2, the object discriminating device is provided with a plurality of templates (for example, for children as templates for discriminating people, for each distance to the object in the image to be discriminated and for each type of the object. (Template for adult) Memorize and detect the distance to the object shown in the image, select the template according to the distance, and then apply the template to perform object discrimination, so pattern matching It is possible to efficiently perform object discrimination by the method even for different types.

  According to the invention of claim 3, the object discriminating apparatus stores a plurality of templates for each distance to the object in the image to be discriminated and the size of the object, and the object displayed in the image The object is discriminated by applying the template after selecting the template according to the distance and size, and then selecting the template according to the distance and size. More efficiently.

  According to the invention of claim 4, this object discriminating apparatus is a group of templates having different modes with respect to an object to be discriminated (for example, a template for discriminating a person who is stopped or a template for discriminating a running person). ) Is stored for each distance to the object in the image to be discriminated, the distance to the object shown in the image is detected, and a group of templates having different modes is selected according to the distance. Since the object discrimination is performed by applying these group of templates, the object discrimination by the pattern matching method can be efficiently performed based on the object mode.

  According to the invention of claim 5, the object discriminating apparatus determines from which region a template is preferentially applied to perform object discrimination when there are a plurality of objects to be discriminated in the image. Based on the specified priority information (for example, priority table), the object is discriminated by applying a template selected according to various factors (for example, the distance to the object and the size of the object) in order. Therefore, it is possible to perform object discrimination by applying a template from a region in the image where the object discrimination should be performed with priority, and the object discrimination can be performed more efficiently.

  According to the invention of claim 6, when there are a plurality of objects to be discriminated in the image, the object discriminating device preferentially applies the template from the object having the smallest distance to the object. Based on the priority information specified to perform object discrimination, templates are preferentially assigned to perform object discrimination, so that it is possible to more efficiently perform object discrimination from near-distance objects that have high object discrimination needs. It becomes possible.

  Further, according to the invention of claim 7, the object discriminating apparatus sequentially gives priority to various factors as described above (for example, the distance to the object) based on the priority information until the next image is input. Since the object determination is performed by applying a template selected in accordance with the size of the object and the size of the object, it is possible to start the object determination for the next image that should be prioritized over the current processing.

  Exemplary embodiments of an object discrimination device, an object discrimination method, and an object discrimination program according to the present invention will be described below in detail with reference to the accompanying drawings.

  In the following first embodiment, the outline and features of the object discriminating apparatus according to the first embodiment, the configuration of the object discriminating apparatus, and the flow of processing will be described in order, and finally the effects of the first embodiment will be described. In the following, an example in which an object in an image captured by an in-vehicle imaging device is determined will be described.

[Outline and Features of Object Discriminating Device (Example 1)]
First, the outline and features of the object discrimination device according to the first embodiment will be described with reference to FIG. FIG. 1 is a diagram for explaining the outline of the object determination device according to the first embodiment. As shown in the figure, the object discriminating apparatus according to the first embodiment applies a template in an image to discriminate a detected object, but from the image to an object detected as a discrimination target. The main feature is that the object is discriminated by detecting the distance and applying the template selected according to the distance, so that the object discrimination can be performed efficiently.

  This main feature will be briefly described. For example, as illustrated in FIG. 1, the object discrimination device stores a plurality of templates for each distance (for example, every 1 m) to an object that can be a discrimination target. . Also, an object to be discriminated is detected from the current frame image (see (1) in FIG. 1). Here, when an object to be discriminated is detected from the current frame image, the background difference method (interframe difference method) or the saliency calculation method is used.

  In the background subtraction method, each pixel in the currently input image is compared with each pixel in the image that was input immediately before, and a difference is extracted. This is a method for detecting an object. According to this, since it is not influenced by a minute pixel change, it is possible to prevent an excessive reaction.

  In addition, the saliency calculation method is a saliency that is not seen in the learned background by learning in advance the partial displacement amount such as brightness, saturation, and line direction of each part of the image as the feature of each part. This is a method for detecting a region where a candidate object is located in an image based on whether or not a special feature (eg, line direction) appears in the current image. According to this, even if the change of the pixel value is small, if there is something unique in the line direction (for example, vector) or the like, a remarkable feature appears and it can be detected as a candidate object.

  Subsequently, when the object discriminating apparatus detects an object to be discriminated from the current frame image by the above-described method, the object discriminating apparatus obtains a distance to the object (see (2) in FIG. 1). Here, the distance to the object is obtained by various ranging methods (stereo method, monocular method, active sensor method), or the vertical position of the object of each pixel in the current frame image (for example, an interval on the image or a pixel) Calculate the approximate distance from (number).

  Then, the object discrimination device selects from a plurality of templates that store templates corresponding to the distance to the object obtained earlier (for example, if the distance to the object is 10 m, the template corresponding to 10 m is selected). Select). Then, this template is applied to an area where an object detected as a discrimination target in the current frame image is located in the image to perform object discrimination (see (3) in FIG. 1).

  As described above, this object discriminating apparatus stores a plurality of templates for each distance to an object in the image to be discriminated, detects the distance to the object shown in the image, and matches the distance to the template. Since the object is discriminated by applying the template after selecting the object, the object discrimination by the pattern matching method can be performed efficiently. Furthermore, since this object discrimination device detects an object in the image and performs pattern matching based on the distance to the detected object, template matching can be performed efficiently even if the road region cannot be specified. Regardless of whether or not is detected, the processing efficiency can be improved as compared with the template matching.

[Configuration of Object Discriminating Device (Example 1)]
Next, the configuration of the object discrimination device 30 whose outline has been described with reference to FIG. 1 will be described with reference to FIG. FIG. 2 is a block diagram illustrating the configuration of the object discrimination device 30 according to the first embodiment. As shown in the figure, the object determination device 30 is connected to a navigation unit 11, a radar unit 12, a preprocessing unit 13, an object detection device 14, and a collision risk determination unit 15. The object discrimination device 30 includes a storage unit 31 and a control unit 32.

  Among them, the navigation (navigation) unit 11 is a means for setting and guiding a travel route from the specified vehicle position and map data by communicating with a GPS (Global Positioning System) artificial satellite. The navigation unit 11 also supplies the driver with various types of information useful for vehicle driving operations such as vehicle position information, road shape, road width, and inclination.

  The radar unit 12 is means for irradiating a radar (for example, 76.5 GHz millimeter wave) and measuring a distance, speed, direction, and the like with a vehicle or an obstacle ahead. The preprocessing unit 13 is a processing unit that performs preprocessing on an image transmitted from the imaging device 10 (for example, a visible light camera, a monocular camera, or a stereo camera), and includes a filter unit 13a and a contour extraction unit 13b. The Among these, the filter unit 13a is means for performing preprocessing (for example, sharpness, contrast adjustment, and saturation adjustment) for enhancing the outline of an object displayed in the image. The contour extracting unit 13b is means for extracting the contour of the object in the image based on the filtering performed by the filter unit 13a.

  The object detection device 14 is a device that detects an object from an image received from the imaging device 10 via the preprocessing unit 13 (restricts an area in the image). For example, an area occupied by an object projected in an image is detected by image processing using a background difference method (interframe difference method) or a saliency calculation method (a region to which a template is applied is narrowed down). And the object discrimination device 30 mentioned later performs image processing using the template matching method for this region. Note that the processing result by the object detection device 14 is transmitted to the object determination device 30 and the collision risk determination unit 15.

  The collision risk determination unit 15 is a processing unit that predicts and determines the collision risk with the object detected from the image based on the processing results received from the object detection device 14 and the object determination device 30. For example, when the object detected by the object determination device 30 is determined to be a pedestrian and is moving near the vehicle from the position in the image, or when the determination result by the object determination device 30 is a child, a collision occurs. It is determined that the risk level is high, and it is determined that the collision risk level is low when the vehicle is stationary at a position away from the vehicle. The processing result by the collision risk determination unit 15 is transmitted to the vehicle control unit 16 and the notification control unit 17.

  Among these, the vehicle control unit 16 is a processing unit that performs vehicle control based on the processing result received from the collision risk determination processing unit 15. For example, when a determination result that the risk of collision with an object detected by the object detection device 14 is high is received, control for stopping the vehicle is performed (for example, automatically even if the driver does not step on the brake). To stop the vehicle safely by intervening in vehicle control). Further, when the determination result that the risk of collision with the object detected by the object detection device 14 is low is received, the control for decelerating the vehicle or the assist of the vehicle control (for example, reducing the play of the brake, Prepare to brake the vehicle).

  The notification control unit 17 is a processing unit that notifies a warning based on the processing result received from the collision risk determination processing unit 15. For example, when a determination result that the risk of collision with an object detected by the object detection device 14 is high is received, a warning is notified from the in-vehicle notification unit 18 by an alarm sound or voice. Further, when a determination result indicating that the risk of collision with an object detected by the object detection device 14 is low, attention is urged from the in-vehicle notification unit 18 by voice or image.

  The in-vehicle notification unit 18 is a means for notifying information from the navigation unit 11 and the notification control unit 17 and includes a display, a speaker, and the like. For example, a part of the display blinks to alert the driver, or a warning or warning is given to the driver by sending a message or an alarm sound from a speaker.

  The storage unit 31 of the object discrimination device 30 is a storage unit (storage unit) that stores data and programs necessary for various processes performed by the control unit 32. In particular, the template storage unit 31a is closely related to the present invention. Is provided.

  The template storage unit 31a is a means for storing various types of information related to the template applied to the object detected by the object detection device 14 from within the image. Specifically, as illustrated in FIG. A plurality of templates are stored for each distance to the object detected (narrowed down) as the target of (i.e., the distance from the front end of the host vehicle to the object). In FIG. 3, the template storage unit 31a is configured to store templates every 1 m. However, the template storage unit 31a is not limited to each 1 m and may be configured to store templates every several m or several cm. .

  The control unit 32 of the object discrimination device 30 is a processing unit that has a predetermined control program, a program that defines various processing procedures, and an internal memory for storing required data, and executes various processes using these. Particularly, those closely related to the present invention include a template selection unit 32a and an object determination unit 32b.

  Of these, the template selection unit 32a is a processing unit that selects a template from the template storage unit 31a. Specifically, the object selection unit 14 measures various distances from the image to the object detected as a discrimination target. A template corresponding to the distance is selected from the template storage unit 31a based on the distance method (stereo method, monocular method, active sensor) or prediction from the vertical position of the object in the image.

  In this case, for example, the distance from the actual front end of the vehicle to the object may be obtained based on the interval on the image from the lower end of the image to the object in the image, or a radar (for example, a millimeter wave radar). Thus, the distance from the front end of the host vehicle to the object may be obtained.

  The object discriminating unit 32b is a processing unit that discriminates what the object is by applying a template to an object detected as a discrimination target from within the image. The template selected from the template storage unit 31a by the template selection unit 32a is applied to the object detected as the discrimination target. And it is discriminate | determined from the presence or absence of matching of a template and an object whether the object is an object shown in a template (for example, discrimination | determination is based on whether a collation degree is higher or lower than a predetermined threshold value). At that time, if the collation degree is lower than a predetermined threshold, it is judged that there is no match, and the templates with the next highest priority are collated in order by applying a template with the next lowest priority.

[Processing by Object Discriminating Device (Example 1)]
Next, processing performed by the object discrimination device according to the first embodiment will be described with reference to FIG. FIG. 4 is a flowchart illustrating a process flow of the object determination device according to the first embodiment. As shown in the figure, upon receiving an image input (Yes at step S401), the object detection device 14 detects an object from the current frame image (step S402).

  As a result, when an object to be discriminated is detected from the current frame image (Yes at Step S403), the template selection unit 32a obtains a distance to the object (Step S404). Here, the distance to the object is obtained by various ranging methods (stereo method, monocular method, active sensor method), or is obtained by predicting an approximate distance from the vertical position of the object in the current frame image.

  Subsequently, the template selection unit 32a selects a template corresponding to the distance to the object to be determined from the template storage unit 31a (step S405). Then, this template is sent to the object discriminating unit 32b.

  The object discriminating unit 32b that has received the template from the template selection unit 32a assigns this template to the object to be discriminated detected from the current frame image, and executes the object discriminating process (step S406).

  Here, returning to the description of step S403, if the object detection device 14 does not detect an object to be discriminated from within the current frame image (No at step S403), it waits for the next image to be input.

  Note that the processing described above is repeated as long as an image is input. If there is no image input, the object discrimination device 30 ends the processing. When the next image is input during the object discrimination process (template fitting), the object discrimination process is stopped and the process is started for the next image.

[Effect of Example 1]
As described above, according to the first embodiment, the object discriminating device 30 stores a plurality of templates for each distance to an object in the image to be discriminated, and the distance to the object displayed in the image. Is detected, and a template matching the distance is selected, and then object identification is performed by applying the template. Therefore, it is possible to efficiently perform object identification using the pattern matching method. Furthermore, since this object discrimination device detects an object in the image and performs pattern matching based on the distance to the detected object, template matching can be performed efficiently even if the road region cannot be specified. Regardless of whether or not is detected, the processing efficiency can be improved as compared with the template matching.

  Further, according to the first embodiment, the object discrimination device selects according to various factors (for example, the distance to the object and the size of the object) based on the priority table until the next image is input. Since the object discrimination is performed by sequentially applying the templates that have been performed, it is possible to start the object discrimination for the next image that should be prioritized over the current processing.

  The object discrimination device according to the first embodiment has been described so far, but the present invention may be implemented in various different forms other than the above-described embodiments. Therefore, in the following, as the second embodiment, various different embodiments are divided into (1) to (6) and described.

(1) Template type In the above embodiment, a plurality of templates are stored for each distance to the object in the image, the distance to the object is detected, and the object selected by applying the template selected according to the distance is used for object discrimination. Although the case where it performs is demonstrated, this invention is not limited to this, The several template is memorize | stored for every distance and classification to the object in an image, the distance to an object is detected, and it matched with the distance It is also possible to select templates by type and perform object discrimination by applying these templates. Hereinafter, the object discrimination device in this case will be described.

  Specifically, as illustrated in FIG. 5, the object discrimination device stores a plurality of templates for each distance and type (for children and adults) to an object in the image. Then, this object discrimination device detects the distance (for example, 10 m) to the object in the image, selects the template for children and adults according to the distance, and both the objects detected as the targets of discrimination. The object is identified by applying the template. As a result, it is possible to efficiently perform object discrimination using a pattern matching method even for different types.

(2) Template according to the size of the object In the above embodiment, a plurality of templates are stored for each distance to the object in the image, the distance to the object is detected, and the template selected according to the distance is selected. Although the case where object discrimination is performed by applying is described, the present invention is not limited to this, and a plurality of templates are stored for each distance to the object and the size of the object in the image, and the distance to the object and the object are stored. It is also possible to detect the size of the object and apply the template selected according to the distance and the size to the object detected as a determination target to perform the object determination. Hereinafter, the object discrimination device in this case will be described.

  Specifically, as illustrated in FIG. 6, the object determination apparatus stores a plurality of templates for each distance to the object and the size of the object (for example, 160 cm or 180 cm) in the image. And this object discrimination device is the distance (for example, 10 m) and the size (for example, 180 cm, the distance to the object in the image and the vertical image of the object imaged in the image. The template selection unit calculates from the upper interval or the number of pixels), selects a template according to the distance and size, and applies this template to the object detected as a discrimination target to perform object discrimination. . This makes it possible to more efficiently perform object discrimination using the pattern matching method.

(3) Aspect of template In the above embodiment, a plurality of templates are stored for each distance to the object in the image, the distance to the object is detected, and the object discriminated by applying the template selected according to the distance. However, the present invention is not limited to this. A group of templates having different modes is stored for each distance to an object in the image, and the distance to the object in the image is detected. Object discrimination may be performed by applying a group of templates having different modes selected according to the distance. The object discrimination device in this case will be described below.

  Specifically, as illustrated in FIG. 7, this object discrimination device is a group of templates having different aspects for each distance to an object in an image (for example, a template for discriminating a standing person or a running person). For example, a template for determining the And this object discrimination device detects the distance (for example, 10 m) to the inside of the image, selects a group of templates having different modes in accordance with the distance, and sets the group of objects detected as a discrimination target to these groups. Object identification is performed by applying a template. As a result, it is possible to efficiently perform the object discrimination by the pattern matching method based on the object mode.

  In the above (1) to (3), when storing a template for each distance to the object and each type of object, when storing a template for each distance to the object and the size of the object, However, the present invention is not limited to this, and more specifically, the object discriminating apparatus, as illustrated in FIG. The template may be stored according to the distance to the object, the type of the object, the size of the object, and the form of the object. Thereby, it becomes possible to perform object discrimination more efficiently based on various situations.

(4) Precedence control for fitting a template In the above embodiment, when an object is detected from an image, a template selected according to the distance to the object is applied to perform object discrimination. However, the present invention is not limited to this, and when a plurality of objects to be discriminated are detected from within the image, it may be controlled from which object the discrimination is performed preferentially. Hereinafter, the object discrimination device in this case will be described.

  Specifically, as illustrated in FIG. 9, the object discrimination device detects an object to be discriminated from the current frame image (see (1) in FIG. 9). Subsequently, the object discriminating apparatus obtains the distance to the detected object from the image and performs labeling in the order of the closest distance to the object (see (2) in FIG. 9). Here, the labeling is performed based on a priority table as illustrated in FIG. This priority table prescribes from which region a template is applied in priority when there are a plurality of objects to be discriminated in the image. For example, the order in which templates are applied (from the shortest to the longest) Is stored and configured. Then, this object discrimination device selects a template corresponding to the distance to the object in the image, and applies the template to the object in the image in the order of labeling to perform object discrimination (see (3) in FIG. 9).

  In this way, various factors (for example, distance to an object and object) are sequentially preferentially based on a priority table that prescribes from which region in which an object is detected a template is preferentially applied to perform object discrimination. Since the object is determined by applying the template selected according to the size of the object), the object can be determined by applying the template from the area in the image where the object should be determined with priority. Can be performed automatically. In addition, based on the priority table that preferentially applies the template from the region where the distance to the object in the image is the smallest, object identification is performed by sequentially applying the template. It becomes possible to perform object discrimination more efficiently from a close object with high necessity.

(5) Device Configuration Each component of the object discrimination device 30 shown in FIG. 2 is functionally conceptual and does not necessarily need to be physically configured as illustrated. That is, the specific form of dispersion / integration of the object discriminating apparatus 30 is not limited to the one shown in the figure. For example, the template selecting section 32a and the object discriminating section 32b are integrated, or the object detecting apparatus 14 and the object discriminating apparatus 30 are integrated. For example, all or a part thereof can be configured to be functionally or physically distributed / integrated in arbitrary units according to various loads or usage conditions. Further, all or a part of each processing function performed in the object discrimination device 30 is realized by a CPU and a program analyzed and executed by the CPU, or is realized as hardware by wired logic. obtain.

(6) Object Discriminating Program By the way, various processes described in the first embodiment are realized by executing a program prepared in advance on a computer system such as a personal computer or a workstation (for example, a vehicle-mounted computer). can do. Therefore, in the following, an example of a computer that executes an object determination program having the same function as in the first embodiment will be described with reference to FIG. FIG. 11 is a diagram illustrating a computer that executes an object determination program.

  As shown in the figure, a computer 40 as an object discrimination device is configured by connecting an HDD 41, a RAM 42, a ROM 43, and a CPU 44 via a bus 50.

  The ROM 43 stores in advance an object discrimination program that exhibits the same function as the object discrimination device 30 shown in the first embodiment, that is, a template selection program 43a and an object discrimination program 43b as shown in FIG. Has been. Note that these programs 43a and 43b may be appropriately integrated or distributed in the same manner as each component of the object determination device 30 shown in FIG. The ROM 43 may be a nonvolatile “RAM”.

  Then, the CPU 44 reads out these programs 43a and 43b from the ROM 43 and executes them, so that each program 43a and 43b functions as a template selection process 44a and an object discrimination process 44b as shown in FIG. . Each process 44a and 44b corresponds to the template selection unit 32a and the object determination unit 32b shown in FIG. 2, respectively.

  The HDD 41 is provided with a template table 41a as shown in FIG. This template table 41a corresponds to the template storage unit 31a shown in FIG. Then, the CPU 44 reads the template data 42a from the template table 41a, stores it in the RAM 42, and executes an object discrimination process based on the template data 42a stored in the RAM 42.

  The above-described programs 43a and 43b do not necessarily need to be stored in the ROM 43 from the beginning. Via a “portable physical medium” such as a magnetic disk or an IC card, or a “fixed physical medium” such as an HDD provided inside or outside the computer 40, or via a public line, the Internet, a LAN, a WAN, etc. Each program may be stored in “another computer (or server)” connected to the computer 40, and the computer 40 may read and execute each program therefrom.

  In the above-described embodiment, the case where an object in an image captured by an in-vehicle imaging device is determined has been described. However, the present invention is not limited to this, and a fixed imaging device (for example, an intersection) You may make it discriminate | determine the object imaged with the traffic flow monitoring camera attached to (3).

  As described above, the object discriminating apparatus, the object discriminating method, and the object discriminating program according to the present invention are useful when a template is applied to a predetermined area in an image and an object in the area is discriminated. It is suitable for efficiently performing object discrimination by a matching method.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a diagram for explaining an overview of an object determination device according to a first embodiment. 1 is a block diagram illustrating a configuration of an object discrimination device according to Embodiment 1. FIG. 6 is a diagram illustrating a configuration example of a template according to Embodiment 1. FIG. 3 is a flowchart illustrating a processing flow of the object determination device according to the first embodiment. It is a figure which shows the structural example of the memorize | stored template which considered the classification of the object which concerns on Example 2. FIG. It is a figure which shows the memorize | stored template structural example which considered the magnitude | size of the object which concerns on Example 2. FIG. It is a figure which shows the structural example of the memorize | stored template which considered the aspect of the object which concerns on Example 2. FIG. It is a figure which shows the structural example of the memorize | stored template which considered the various factors based on Example 2. FIG. FIG. 10 is a diagram for explaining an outline in the case of controlling the order in which templates are applied to a plurality of regions to be determined according to the second embodiment. It is a figure which shows the structural example of the priority table which concerns on Example 2. FIG. It is a figure which shows the computer which performs an object discrimination | determination program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Imaging device 11 Navigation (navigation) part 12 Radar part 13 Preprocessing part 13a Filter part 13b Contour extraction part 14 Object detection apparatus 15 Collision risk determination part 16 Vehicle control part 17 Notification control part 18 In-car notification part 30 Object discrimination apparatus 31 Storage unit 31a Template storage unit 32 Control unit 32a Template selection unit 32b Object discrimination unit 40 Computer 41 HDD (Hard disk drive)
42 RAM (Random Access Memory)
43 ROM (Read Only Memory)
44 CPU (Central Processing Unit)
50 buses

Claims (9)

In an object discriminating apparatus that applies a template in an image and discriminates a detected object,
Template storage means for storing a plurality of templates for each distance to the object to be determined;
A template selection means for detecting a distance to the object from the templates stored in the template storage means and selecting a template corresponding to the distance;
Object discriminating means for discriminating the object by applying the template selected by the template selecting means in the image;
An object discrimination device comprising: The template storage means stores a plurality of templates for each distance to the object to be determined and the type of the object,
The template selection means detects the distance to the object in the image from the templates stored in the template storage means, and selects a template corresponding to the distance for each type of the object,
2. The object discriminating apparatus according to claim 1, wherein the object discriminating unit discriminates the object by applying various types of templates selected by the template selecting unit to the image. The template storage means stores a plurality of templates for each distance to the object to be determined and the size of the object,
The template selection means detects the distance and size to the object from the templates stored in the template storage means, selects a template corresponding to the distance and size,
The object discriminating apparatus according to claim 1, wherein the object discriminating unit discriminates the object by applying a template selected by the template selecting unit to the image. The template storage means stores, for each predetermined distance, a group of templates having different aspects with respect to the object to be determined.
The template selection means detects a distance to the object from the templates stored in the template storage means, selects a different group of templates of the aspect corresponding to the distance,
4. The object discriminating apparatus according to claim 2, wherein the object discriminating unit discriminates the object by applying a group of templates having different modes selected by the template selecting unit to the image. . Priority information defining which object is to be preferentially applied to the template when there are a plurality of objects to be discriminated in the image,
Priority control means for controlling the templates selected by the template selection means to be preferentially applied within the image based on the priority information;
The object discriminating apparatus according to claim 1, further comprising: The priority information stipulates that the template is preferentially applied sequentially from the object having the smallest distance to the object to be determined,
The priority control means controls the templates selected by the template selection means so as to be preferentially assigned based on the priority information from the object having the smallest distance to the object to be determined. The object discriminating apparatus according to claim 5.   The priority control unit controls the template selected by the template selection unit to be applied to a predetermined region in the image based on the priority information until a next image is input. The object discrimination device according to claim 5. In an object discrimination method for applying a template in an image and discriminating a detected object,
A template storage step of storing a plurality of templates for each distance to the object to be determined;
A template selection step of detecting a distance to the object from the templates stored in the template storage means and selecting a template corresponding to the distance;
An object discriminating step of discriminating the object by applying the template selected by the template selecting means in the image;
An object discrimination method characterized by comprising: In an object discrimination program for applying a template in an image and causing a computer to execute a method for discriminating a detected object,
A template storage procedure for storing a plurality of templates for each distance to the object to be determined;
A template selection procedure for detecting the distance to the object from the templates stored in the template storage means and selecting a template corresponding to the distance;
An object discriminating procedure for discriminating the object by applying the template selected by the template selecting means in the image;
An object discrimination program for causing a computer to execute.

Images