Image processing device, computer readable storage medium, and image processing method

Abstract

The present invention provides an image processing device, a computer-readable storage medium, and an image processing method, which can efficiently set the values ​​of multiple parameters. The image processing device (50) processes a processing image based on multiple parameters, and the image processing device (50) includes: a prior knowledge generation unit (44) that uses multiple learning images, evaluation items corresponding to the multiple learning images, and target values ​​of the evaluation items to generate prior knowledge (34) related to the parameters; and a setting unit (42) that sets the values ​​of the parameters based on the prior knowledge (34), multiple processing images, evaluation items corresponding to the multiple processing images, and target values ​​of the evaluation items.

Description

Technical Field

[0001] The present invention relates to an image processing device, a computer-readable storage medium and an image processing method. Background Art

[0002] Conventionally, as a parameter determination support device, there is known a parameter determination support device as follows, that is, a parameter determination support device for an image processing device, the image processing device performs processing using a predetermined parameter group on image data obtained by photographing a measurement target object to obtain a processing result, the parameter determination support device comprising: an input unit that receives image data and an expected result associated with the image data; a candidate generation unit that generates a plurality of parameter candidates in which at least one parameter value included in the parameter group is different from each other; an acquisition unit that acquires a plurality of processing results using the plurality of parameter candidates for the image data; an evaluation unit that compares each of the plurality of processing results with the corresponding expected result to generate evaluation results for each of the plurality of processing results; and an output unit that outputs the evaluation results according to the plurality of parameter candidates (see Patent Document 1). The parameter determination support device can more quickly and easily determine a parameter group to be set for the image processing device, the image processing device performs processing using a predetermined parameter group on image data obtained by photographing a measurement target object to obtain a processing result.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Application Publication No. 2010-191939 Summary of the invention

[0006] Problems to be solved by the invention

[0007] On the other hand, when setting values ​​of multiple parameters, in order to set appropriate parameter values, for example, all combinations of values ​​of the multiple parameters must be evaluated and the combination of parameter values ​​with the highest evaluation value must be selected. Therefore, setting the parameter values ​​takes a very long time and is inefficient.

[0008] Therefore, an object of the present invention is to provide an image processing device, an image processing program, and an image processing method capable of efficiently setting values ​​of a plurality of parameters.

[0009] Technical means of solving problems

[0010] An image processing device according to one embodiment of the present invention processes a processing image based on multiple parameters, and the image processing device includes: a prior knowledge generating unit, which generates prior knowledge related to the parameters using multiple learning images, evaluation items corresponding to the multiple learning images, and target values ​​of the evaluation items; and a setting unit, which sets the values ​​of the parameters based on the prior knowledge, multiple processing images, evaluation items corresponding to the multiple processing images, and target values ​​of the evaluation items.

[0011] According to this embodiment, the value of the parameter is set based on prior knowledge related to the parameter, a plurality of processing images, evaluation items corresponding to the plurality of processing images, and target values ​​of the evaluation items. Thus, for example, for a certain parameter, an appropriate range of values ​​can be explored with an appropriate change range, so that the appropriate parameter value can be set in a short time compared to a case without prior knowledge. Therefore, the values ​​of a plurality of parameters can be set efficiently.

[0012] In the aforementioned embodiment, the prior knowledge generating unit may also generate prior knowledge based on an evaluation function determined according to an evaluation item and a target value of the evaluation item, an evaluation component for obtaining an evaluation value of the evaluation function corresponding to a parameter value, and a plurality of evaluation values ​​obtained using the evaluation component, and the setting unit sets the value of the parameter in such a way that the evaluation value becomes higher.

[0013] According to this embodiment, the value of the parameter is set in a manner that makes the evaluation value higher. Thus, it is possible to set the value of the parameter close to or satisfying the target value of the evaluation item. Moreover, using an evaluation function determined according to the evaluation item and the target value of the evaluation item, prior knowledge is generated based on an evaluation component for obtaining the evaluation value of the evaluation function corresponding to the value of the parameter and a plurality of evaluation values ​​obtained using the evaluation component. Thus, for example, for a parameter suitable for improving the evaluation value, it is possible to explore the value of the appropriate range with an appropriate change range, so compared with a situation without prior knowledge, the value of the parameter with a higher evaluation value can be set in a short time.

[0014] In the aforementioned embodiment, the evaluation component may also be an evaluation value map, which is generated using an evaluation function and represents the evaluation value of each parameter value in the evaluation item.

[0015] According to this embodiment, the evaluation means generates an evaluation value map indicating the evaluation value of each parameter value in the evaluation item using the evaluation function. Thus, the prior knowledge generation unit that generates prior knowledge can be easily realized.

[0016] In the aforementioned embodiment, the prior knowledge generating unit may generate prior knowledge for each type of processing or each variety of an object of processing an image, and the image processing device may further include a selecting unit that selects prior knowledge based on the specified type or variety.

[0017] According to this embodiment, prior knowledge is generated for each type of processing or each type of object to be processed, and prior knowledge is selected based on the designated type or type. Thus, prior knowledge common to each type of processing or each type of object can be obtained.

[0018] In the aforementioned embodiment, the prior knowledge includes at least one of the influence of the parameter on the evaluation item, the dependency between the parameters, the variation range of the parameter value, and the variation range of the parameter value.

[0019] According to this embodiment, the prior knowledge includes at least one of the influence of the parameter on the evaluation item, the dependency between the parameters, the change range of the parameter value, and the change range of the parameter value. Thus, it is possible to easily realize a setting unit that sets a parameter with a higher evaluation value in a short time.

[0020] In the above-mentioned embodiment, the setting unit may set the value having the highest evaluation value within the specified setting time range as the parameter.

[0021] According to this embodiment, the value with the highest evaluation value within the specified setting time range is set as the parameter. Thus, even within the limited setting time, the value of the parameter with a higher evaluation value can be set.

[0022] The aforementioned embodiment may also include: an extraction unit that extracts at least one of the plurality of parameters based on prior knowledge, and a setting unit that sets a value of the extracted parameter.

[0023] According to this embodiment, at least one of the plurality of parameters is extracted based on prior knowledge, and the value of the extracted parameter is set. Thus, the value of the parameter that can improve the evaluation value among the plurality of parameters can be set.

[0024] Moreover, an image processing program of another embodiment of the present invention is executed by a computer, and processes a processing image based on multiple parameters. The image processing program includes: a prior knowledge generation step, using multiple learning images, evaluation items corresponding to the multiple learning images, and target values ​​of the evaluation items to generate prior knowledge related to the parameters; and a setting step, setting the value of the parameter based on the prior knowledge, the multiple processing images, the evaluation items corresponding to the multiple processing images, and the target values ​​of the evaluation items.

[0025] According to this embodiment, the value of the parameter is set based on prior knowledge related to the parameter, a plurality of processing images, evaluation items corresponding to the plurality of processing images, and target values ​​of the evaluation items. Thus, for example, for a certain parameter, an appropriate range of values ​​can be explored with an appropriate change range, so that the appropriate parameter value can be set in a short time compared to a case without prior knowledge. Therefore, the values ​​of a plurality of parameters can be set efficiently.

[0026] In the aforementioned embodiment, the prior knowledge generating step may also include: using an evaluation function determined according to an evaluation item and a target value of the evaluation item, generating the prior knowledge based on an evaluation component for obtaining an evaluation value of the evaluation function corresponding to a parameter value, and a plurality of evaluation values ​​obtained using the evaluation component, and the setting step includes: setting the value of the parameter in such a way that the evaluation value becomes higher.

[0027] According to this embodiment, the value of the parameter is set in a manner that makes the evaluation value higher. Thus, it is possible to set the value of the parameter close to or satisfying the target value of the evaluation item. Moreover, using an evaluation function determined according to the evaluation item and the target value of the evaluation item, prior knowledge is generated based on an evaluation component for obtaining the evaluation value of the evaluation function corresponding to the value of the parameter and a plurality of evaluation values ​​obtained using the evaluation component. Thus, for example, for a parameter suitable for improving the evaluation value, it is possible to explore the value of the appropriate range with an appropriate change range, so compared with a situation without prior knowledge, the value of the parameter with a higher evaluation value can be set in a short time.

[0028] In the aforementioned embodiment, the evaluation component may also be an evaluation value map, which is generated using an evaluation function and represents the evaluation value of each parameter value in the evaluation item.

[0029] According to this embodiment, the evaluation means generates an evaluation value map indicating the evaluation value of each parameter value in the evaluation item using the evaluation function. Thus, the prior knowledge generation step of generating prior knowledge can be easily realized.

[0030] In the aforementioned embodiment, the prior knowledge generation step may also include: generating prior knowledge for each type of processing or each variety of object of the image to be processed, and the image processing program also includes: a selection step of selecting prior knowledge based on the specified type or variety.

[0031] According to this embodiment, prior knowledge is generated for each type of processing or each type of object of the processing image, and an evaluation value map is selected for each type of processing or each type of object. Thus, prior knowledge common to each type of processing or each type of object can be obtained.

[0032] In the aforementioned embodiment, the prior knowledge may also include at least one of the influence of the parameter on the evaluation item, the dependency between the parameters, the change range of the parameter value, and the change range of the parameter value.

[0033] According to this embodiment, the prior knowledge includes at least one of the influence of the parameter on the evaluation item, the dependency between the parameters, the change range of the parameter value, and the change range of the parameter value. Thus, it is possible to easily realize a setting unit that sets a parameter with a higher evaluation value in a short time.

[0034] In the aforementioned embodiment, the setting step may also include: setting the value with the highest evaluation value within the specified setting time range as the parameter.

[0035] According to this embodiment, the value with the highest evaluation value within the specified setting time range is set as the parameter. Thus, even within the limited setting time, the value of the parameter with a higher evaluation value can be set.

[0036] The aforementioned embodiment may also include: an extraction step of extracting at least one of the plurality of parameters based on prior knowledge, and a setting step of setting the value of the extracted parameter.

[0037] According to this embodiment, at least one of the plurality of parameters is extracted based on prior knowledge, and the value of the extracted parameter is set. Thus, the value of the parameter that can improve the evaluation value among the plurality of parameters can be set.

[0038] Moreover, an image processing method according to another embodiment of the present invention processes a processing image based on multiple parameters, and the image processing method includes: a prior knowledge generating step, using multiple learning images, evaluation items corresponding to the multiple learning images, and target values ​​of the evaluation items to generate prior knowledge related to the parameters; and a setting step, setting the value of the parameter based on the prior knowledge, the multiple processing images, the evaluation items corresponding to the multiple processing images, and the target values ​​of the evaluation items.

[0039] According to this embodiment, the value of the parameter is set based on prior knowledge related to the parameter, a plurality of processing images, evaluation items corresponding to the plurality of processing images, and target values ​​of the evaluation items. Thus, for example, for a certain parameter, an appropriate range of values ​​can be explored with an appropriate change range, so that the appropriate parameter value can be set in a short time compared to a case without prior knowledge. Therefore, the values ​​of a plurality of parameters can be set efficiently.

[0040] Effects of the Invention

[0041] According to the present invention, the values ​​of a plurality of parameters can be set efficiently. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a structural diagram illustrating a schematic structure of an image processing system according to an embodiment.

[0043] Figure 2 is an example Figure 1 Plot of the evaluation function shown.

[0044] Figure 3 is an example Figure 1 A diagram of the evaluation value mapping shown.

[0045] Figure 4 is an example Figure 1 A diagram of the evaluation value mapping shown.

[0046] Figure 5 is an example Figure 1 A diagram of the evaluation value mapping shown.

[0047] Figure 6 is an example Figure 1 Figure 2. The learning dataset shown.

[0048] Figure 7 is an example Figure 1 A diagram of the evaluation value mapping shown.

[0049] Figure 8 This is a flowchart showing a schematic operation of setting parameters of the image processing device according to one embodiment.

[0050] [Explanation of Symbols]

[0051] 10: Ministry of Communications

[0052] 20: Input / output I / F

[0053] 30: Storage

[0054] 31: Evaluation Function

[0055] 32: Evaluation value mapping

[0056] 33: Learning Dataset

[0057] 34: Prior Knowledge

[0058] 35: Exploring History

[0059] 36: Calculation results

[0060] 40: Control Department

[0061] 41: Selection Department

[0062] 42: Setting Department

[0063] 43: Extraction Department

[0064] 44: Prior knowledge generation department

[0065] 50: Image processing device

[0066] 51: Bus

[0067] 60: Shooting device

[0068] 70: Input device

[0069] 80: Output device

[0070] 90: Line device

[0071] 100: Image processing system

[0072] NW: Network

[0073] p1, p2, p3: parameters

[0074] S200: Parameter setting process

[0075] S201~S209: Steps

[0076] SV: Server DETAILED DESCRIPTION

[0077] The following describes the embodiments of the present invention. In the following description of the drawings, the same or similar parts are represented by the same or similar symbols. However, the drawings are schematic. Therefore, specific dimensions and the like should be judged in accordance with the following description. Moreover, the drawings also include parts with different dimensional relationships or ratios. Furthermore, the technical scope of the present invention should not be limited to the embodiments described.

[0078] First, refer to Figures 1 to 7 An example of the structure of an image processing device according to an embodiment of the present invention will be described. Figure 1 1 is a block diagram illustrating a schematic structure of an image processing system 100 according to an embodiment. Figure 2 is an example Figure 1 A diagram of the evaluation function 31 is shown. Figure 3 is an example Figure 1 A diagram of an evaluation value map 32 is shown. Figure 4 is an example Figure 1 A diagram of an evaluation value map 32 is shown. Figure 5 is an example Figure 1 A diagram of an evaluation value map 32 is shown. Figure 6 is an example Figure 1 A diagram of a learning dataset 33 is shown. Figure 7 is an example Figure 1 A diagram of an evaluation value map 32 is shown.

[0079] like Figure 1 As shown, the image processing system 100 includes a server SV, a network NW, an image processing device 50 , a photographing device 60 , an input device 70 , an output device 80 , and a line device 90 .

[0080] The image processing device 50 and the server SV are connected to each other via a network NW so as to be able to communicate with each other. The network NW is, for example, the Internet, a local area network (LAN), a dedicated line, a telephone line, an intranet, Bluetooth (registered trademark), Wireless Fidelity (Wi-Fi), a mobile communication network, other communication lines, or a combination thereof. Furthermore, the connection between the image processing device 50 and the server SV and the network NW may be wired or wireless.

[0081] In this embodiment, Figure 1 2 shows an example in which the image processing system 100 includes one network NW, but the present invention is not limited to this. For example, the image processing system 100 may include two or more networks.

[0082] The server SV is a computer that functions as a higher-level controller of the image processing device 50. The server SV is configured to include a controller such as a programmable logic controller (PLC), for example.

[0083] The image processing device 50 processes the processing image based on a plurality of parameters. More specifically, the image processing device 50 checks the processing image on which the object is reflected based on a plurality of parameters, determines the position based on the processing image on which the object is reflected, or compares the characters based on the processing image on which the object is reflected.

[0084] In this embodiment, for example, the image processing device 50 checks the imaged processing image of the object by photographing the image of the object on the line device 90 through the photographing device 60. At this time, the plurality of parameters are, for example, the size of the kernel function or filter when smoothing the processing image, the threshold value when binarizing the processing image, and the like.

[0085] The image processing device 50 is, for example, an information processing device such as a computer. The image processing device 50 includes a communication unit 10, an input / output interface (hereinafter referred to as "input / output I / F") 20, a storage unit 30, and a control unit 40. In addition, the image processing device 50 also includes a bus 51, and the bus 51 is configured to transmit signals or data between the various units of the image processing device 50.

[0086] The communication unit 10 is used to communicate (transmit and receive) data via the network NW. The communication unit 10 is configured to be able to communicate with a server SV connected to the network NW via the network NW based on one or more predetermined communication methods.

[0087] The input / output I / F 20 is an interface between the image processing device 50 and an external device. The input / output I / F 20 is configured to exchange data or signals with an external device. Moreover, the input / output I / F 20 is configured to control communication with an external device. The input / output I / F 20 is configured to include connection terminals of standardized interfaces such as Universal Serial Bus (USB), High Definition Multimedia Interface (HDMI (registered trademark)), Institute of Electrical and Electronics Engineers (IEEE) 1394, etc. The input / output I / F 20 is connected to the shooting device 60, the input device 70, the output device 80, and the line device 90.

[0088] The storage unit 30 is configured to store programs, data, etc. The storage unit 30 is configured to include, for example, a hard disk drive, a solid state drive, etc. The storage unit 30 stores in advance various programs executed by the control unit 40 or data necessary for executing the programs.

[0089] Furthermore, the storage unit 30 stores in advance an evaluation function 31 , an evaluation value map 32 , a learning data set 33 , prior knowledge 34 , an exploration history 35 , and a calculation result 36 .

[0090] The evaluation function 31 is determined based on the evaluation item and the target value of the evaluation item. The evaluation item is an item corresponding to a plurality of images. In this embodiment, the evaluation item corresponds to a plurality of learning images contained in the learning data set 33. Figure 2 As shown, the evaluation items are, for example, "processing time", "pass rate", "miss rate", "position accuracy", "posture accuracy", "dimension accuracy", etc. For each evaluation item, a target value is set. Figure 2 In the example shown, the target value includes two target values: a necessary (must) target value and a desired (want) target value.

[0091] Each target value is input by a user operating the input device 70. The evaluation function 31 normalizes the quantity by the value of the parameter of the image processing device 50 so that when the necessary target value of all evaluation items is satisfied, it is, for example, 60 points or 60%, and when the expected target value of all evaluation items is satisfied, it is, for example, 90 points or 90%.

[0092] The evaluation value map 32 is used Figure 2 The evaluation function 31 is generated for each evaluation item of the evaluation function 31. The evaluation value map 32 indicates the evaluation value of the evaluation function (hereinafter referred to as "evaluation value") for each parameter value of the image processing device 50 in the evaluation item.

[0093] For example, Figure 3 The evaluation value map 32 shown is one in which evaluation values ​​in the parameter space of parameters p1 , p2 , and p3 are mapped to “processing time” which is one of the evaluation items of the evaluation function 31 .

[0094] and, Figure 4 The evaluation value map 32 shown is a map of the evaluation values ​​in the parameter space of the parameters p1 , p2 , and p3 in the “pass rate” which is another evaluation item of the evaluation function 31 .

[0095] In addition, the evaluation value map 32 is not limited to mapping evaluation values ​​in a plurality of parameter spaces for each evaluation item. Figure 5 As shown, in the evaluation item "processing time", the evaluator for the value of the parameter p1 is mapped.

[0096] Prior knowledge 34 is based, for example, on Figures 3 to 5 The prior knowledge 34 is related to the parameters, and includes, for example, the degree of influence of the parameters on the evaluation items, the degree of dependency between the parameters, the step width of the parameters, the range of change of the parameters, and the like. Figure 5 , an example of obtaining the change range of the parameter p1 as the prior knowledge 34 is shown.

[0097] When the influence of the parameters on the evaluation items is obtained as the prior knowledge 34, when the evaluation value of a certain evaluation item is low, the parameter whose value should be changed in order to improve the evaluation value of the evaluation item can be known. Moreover, when the dependency between the parameters is obtained as the prior knowledge 34, when a certain parameter is changed, the parameter that should be changed at the same time can be known. Moreover, when the range of change of the parameters is obtained as the prior knowledge 34, when a higher value of the evaluation value is explored for a certain parameter, the appropriate range of change in the value of the parameter can be known. Furthermore, when the range of change of the parameters is obtained as the prior knowledge 34, when a higher value of the evaluation value is explored for a certain parameter, the appropriate range of change in the value of the parameter can be known.

[0098] In this embodiment, the prior knowledge 34 is shown as an example generated based on the evaluation value map 32, but it is not limited to this. For example, for the questions prepared in advance, a method of conducting some trial experiments to grasp the tendency of parameters is considered. In this case, the prior knowledge can also be obtained based on the tendency of the parameters.

[0099] The evaluation value map 32 is generated using the learning data set 33. Figure 6 As shown, the learning data set 33 includes a plurality of learning data, and the evaluation value map 32 is generated using at least one learning data in the learning data set 33 .

[0100] Each learning data includes, for example, a plurality of learning images, the type of processing performed on the plurality of learning images, the number of learning images, and a label assigned to each learning image. The label is associated information corresponding to the type of processing, that is, an annotation. For example, when the processing type is "inspection", the label is "OK" or "NG", when the processing type is "positioning", the label is the position (x, y) of the xy coordinates based on a certain pixel as the reference or origin, and when the processing type is "character comparison", the label is a date such as year and month.

[0101] The evaluation value map 32 may be generated for each evaluation item, for each type of processing performed on the processing image or for each type of object reflected in the processing image. Figure 7 As shown, the evaluation value map 32 is obtained by mapping the evaluation values ​​in the parameter space of p1, p2, and p3 with respect to the evaluation item "processing time" using the learning data whose processing type is "inspection".

[0102] The evaluation value map 32 stored in the storage unit 30 is not limited to the case where it is generated by the image processing device 50. For example, Figure 1 The evaluation value map is created by the server SV shown in FIG. 1 . At this time, the server SV stores the evaluation function and the learning data set in advance, or obtains the evaluation function and the learning data set from other devices via the network NW. In addition, the evaluation value map generated in the server SV is sent to the image processing device 50 via the network NW and stored in the storage unit 30.

[0103] In addition, the structure and format of the evaluation function 31, the evaluation value map 32, the learning data set 33, the prior knowledge 34, the exploration history 35, and the calculation result 36 are not limited to the above-mentioned examples. For example, the evaluation function 31, the evaluation value map 32, the learning data set 33, the prior knowledge 34, the exploration history 35, and the calculation result 36 can be simple data or a database. Moreover, when at least a part of the evaluation function 31, the evaluation value map 32, the learning data set 33, the prior knowledge 34, the exploration history 35, and the calculation result 36 is a database, normalization can be performed and the group unit of the data can be subdivided.

[0104] return Figure 1 As described above, the control unit 40 is configured to control the operation of each unit of the image processing device 50, such as the communication unit 10, the input / output I / F 20, and the storage unit 30. Moreover, the control unit 40 is configured to realize each function described later by executing a program stored in the storage unit 30. The control unit 40 is, for example, configured to include a processor such as a central processing unit (CPU), a memory such as a read-only memory (ROM), a random access memory (RAM), and a buffer storage device such as a buffer. In addition, the details of the control unit 40 will be described later.

[0105] The photographing device 60 is configured to take an image and record it as data. The photographing device 60 is a digital camera, for example, and is composed of optical system parts such as lenses and electronic system parts such as photographing elements (light receiving elements). In addition, the optical system parts may also include a plurality of lenses. Moreover, the electronic system parts may also include light-emitting devices such as flashes. The photographing device 60 is arranged above the line device 90, photographs the object on the line device 90, and outputs the photographed image. The image output by the photographing device 60 is input to the image processing device 50 via the input / output I / F 20. The control unit 40 performs necessary processing on the image input from the photographing device 60 to generate a file of the image for processing, and stores the generated file of the image for processing in the storage unit 30.

[0106] The input device 70 is configured to input information through the operation of the user. The input device 70 includes, for example, a keyboard, a keypad, a mouse, a trackball, a touch panel, a microphone, etc. For example, when the user operates the keyboard, keypad, mouse, trackball, touch panel, microphone, etc. (including voice operation using a microphone), the input device 70 outputs data or signals corresponding to the operation. The data or signal output by the input device 70 is input to the image processing device 50 via the input / output I / F 20. The control unit 40 generates data based on the data or signal, thereby being able to input information to the image processing device 50.

[0107] The output device 80 is configured to output information and includes, for example, a display device such as a liquid crystal display, an electroluminescence (EL) display, or a plasma display, and an audio device such as a speaker.

[0108] The output device 80 can output information by displaying text such as characters, numbers, and symbols, images, and moving image information on a display device and outputting sound information from a speaker, among information input from the image processing device 50 via the input / output I / F 20 .

[0109] The line device 90 is configured to transport an object. The line device 90 includes a transport member such as a belt conveyor. The line device 90 can move, stop or remove an object on the line device 90 based on a control signal input from the image processing device 50 via the input / output I / F 20.

[0110] The control unit 40 of the image processing device 50 includes, for example, a selection unit 41 , a setting unit 42 , an extraction unit 43 , and a prior knowledge generation unit 44 as its functional configuration.

[0111] The prior knowledge generation unit 44 is configured to generate the aforementioned prior knowledge 34 using a plurality of learning images included in the learning data of the learning data set 33, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items. The prior knowledge generation unit 44 stores the generated prior knowledge 34 in the storage unit 30.

[0112] In more detail, the prior knowledge generation unit 44 is configured to generate prior knowledge 34 using an evaluation function 31 determined based on an evaluation item and a target value of the evaluation item, and based on an evaluation component for obtaining an evaluation value of the evaluation function 31 corresponding to a parameter value and a plurality of evaluation values ​​obtained using the evaluation component.

[0113] The aforementioned evaluation means is, for example, the evaluation value map 32, which is generated using the evaluation function 31 and indicates the evaluation value of each parameter value in the evaluation item. Thus, the prior knowledge generation unit 44 that generates the prior knowledge 34 can be easily realized.

[0114] Furthermore, the prior knowledge generating unit 44 may generate the prior knowledge 34 for each type of processing on the processing image or each type of object of the processing image.

[0115] The selection unit 41 is configured to select the prior knowledge 34. More specifically, the selection unit 41 is configured to select the prior knowledge 34 based on the designated type of processing for the image or the type of the object of the image. The type of processing is, for example, inspection, positioning, character matching, etc. And the type of the object is, for example, automobile, food, etc.

[0116] As described above, prior knowledge 34 is generated for each type of processing or each type of object to be processed using an image, and prior knowledge 34 is selected based on the designated type or type, thereby obtaining prior knowledge 34 common to each type of processing or each type of object.

[0117] In addition, the selection unit 41 may select the prior knowledge 34 based on the image given when selecting the prior knowledge 34. In this case, the selection unit 41 uses a network that has undergone machine learning, etc. to search for a learning image similar to the given image, and finds the learning data of the learning data set 33 that includes the similar learning image. Then, the selection unit 41 selects the prior knowledge 34 obtained using the learning data.

[0118] The setting unit 42 is configured to set the value of the parameter based on the prior knowledge 34, a plurality of processing images, evaluation items corresponding to the plurality of processing images, and the target value of the evaluation item. Thus, for example, for a certain parameter, an appropriate range of values ​​can be searched with an appropriate change range, so that the appropriate parameter value can be set in a short time compared to a case without prior knowledge. Therefore, the values ​​of a plurality of parameters can be set efficiently.

[0119] In more detail, the setting unit 42 is configured to set the value of the parameter in such a way that the evaluation value becomes higher or the evaluation value is high. Thus, the value of the parameter can be set to be close to or satisfy the target value of the evaluation item. Moreover, as described above, the prior knowledge 34 is generated based on the evaluation component for obtaining the evaluation value of the evaluation function corresponding to the value of the parameter and the multiple evaluation values ​​obtained by using the evaluation component, using the evaluation function determined according to the evaluation item and the target value of the evaluation item. Thus, for example, for a parameter suitable for improving the evaluation value, it is possible to explore the value of the appropriate range with an appropriate change range, so that compared with the case where there is no prior knowledge 34, the value of the parameter with a higher evaluation value can be set in a short time.

[0120] Preferably, the prior knowledge 34 includes at least one of the degree of influence of the parameter on the evaluation item, the dependency between parameters, the change range of the parameter value, and the change range of the parameter value. Thus, the setting unit 42 can easily set the parameter with a higher evaluation value in a short time.

[0121] Furthermore, the setting unit 42 is configured to set the value with the highest evaluation value within the specified setting time as the parameter when the setting time is specified. Thus, even within the limited setting time, the value of the parameter with a higher evaluation value can be set.

[0122] The setting time is designated by, for example, a user operating the input device 70. The setting time is designated from, for example, pre-prepared options such as 10 seconds, 1 minute, and 1 day.

[0123] The extracting unit 43 is configured to extract at least one of the plurality of parameters based on the prior knowledge 34. When the extracting unit 43 extracts a parameter, the setting unit 42 sets the value of the extracted parameter. Thus, the value of the parameter that can improve the evaluation value among the plurality of parameters can be set.

[0124] Each function of the control unit 40 can be realized by a program executed by a computer (microprocessor). Therefore, each function included in the control unit 40 can be realized by hardware, software, or a combination of hardware and software, and is not limited to any one case.

[0125] Furthermore, when each function of the control unit 40 is implemented by software or a combination of hardware and software, the processing may be performed using multitasking, multithreading, or both multitasking and multithreading, and is not limited to any one case.

[0126] Next, refer to Figure 8 The operation of the image processing device according to one embodiment will be described. Figure 8This is a flowchart showing a schematic operation of setting parameters of the image processing device 50 according to one embodiment.

[0127] For example, when a user selects a parameter setting through an operation of the user, the control unit 40 of the image processing device 50 executes Figure 8 The parameter setting process S200 is shown.

[0128] In the following description, it is assumed that the prior knowledge 34 is generated in advance and stored in the storage unit 30. That is, it is assumed that the prior knowledge generation step of generating the prior knowledge 34 using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items is performed before the parameter setting process S200. Furthermore, it is assumed that the objects of the processing performed by the image processing device 50 based on the plurality of parameters, i.e., a plurality of processing images, are given in advance, and the type of the processing or the type of the object is specified in advance.

[0129] First, the selection unit 41 derives some of the given plurality of processing images as samples (hereinafter referred to as “sample images”) ( S201 ).

[0130] Next, the selection unit 41 selects the prior knowledge 34 based on the designated type of process or type of object (S202). The designation of the type of process or type of object is performed by, for example, a user operating the input device 70.

[0131] Next, the setting unit 42 generates a rough evaluation value map (S203) based on the prior knowledge 34 selected in step S202 using the sample image derived in step S201. In step S203, the setting unit 42 refers to the prior knowledge 34, such as the variation width and range of parameter values.

[0132] Next, the extraction unit 43 extracts a plurality of combinations of parameters with high evaluation values ​​and the values ​​as parameter candidates based on the rough evaluation value map created in step S203 ( S204 ).

[0133] Next, the setting unit 42 searches for the parameter candidate extracted in step S204 based on the prior knowledge 34 selected in step S202 whether there is a parameter value having a higher evaluation value or a vicinity of the value of the parameter candidate ( S205 ).

[0134] As a specific example of the exploration, the setting unit 42 sets the value of the parameter candidate to the central value and sets the variation range to the minimum value. Furthermore, the setting unit 42 changes the value of the parameter within the variation range of the prior knowledge 34 to obtain the evaluation value. At this time, the setting unit 42 refers to the influence of the parameter on the evaluation item, the dependency between the parameters, etc. in the prior knowledge 34, and changes the value of the parameter with a high influence on the evaluation item with a low evaluation value and the parameter with a high dependency between the parameters to obtain the evaluation value.

[0135] Furthermore, during step S205, the setting unit 42 writes the prior knowledge 34 used in the search, the searched parameters, the search range, the change range during the search, etc., into the search history 35 of the storage unit 30 and stores them. Furthermore, the setting unit 42 writes the parameters, the values, and the evaluation values ​​when the evaluation values ​​are obtained during the search into the calculation results 36 of the storage unit 30 and stores them. Thus, when the conditions of the parameters, the values, etc. are the same, the setting unit 42 refers to the calculation results 36 of the storage unit 30, thereby omitting the calculation of the evaluation value.

[0136] Next, the extraction unit 43 extracts a plurality of combinations of parameters with high evaluation values ​​and the values ​​as promising parameter candidates based on the result of the search in step S205 ( S206 ).

[0137] Next, the setting unit 42 searches for a value of a parameter with a higher evaluation value or a vicinity of the value of the promising parameter candidate extracted in step S206 based on the prior knowledge 34 selected in step S202 ( S207 ).

[0138] The specific example of the exploration is the same as the exploration in step S205. The setting unit 42 sets the value of the expected parameter candidate as the central value and sets the variation range to the minimum value. In addition, the setting unit 42 changes the value of the parameter within the range of the variation range of the value of the parameter in the prior knowledge 34 to obtain the evaluation value. At this time, the setting unit 42 refers to the influence of the parameter on the evaluation item, the dependence between the parameters, etc. in the prior knowledge 34, and changes the value of the parameter with a high influence on the evaluation item with a low evaluation value and the parameter with a high dependence between the parameters to obtain the evaluation value.

[0139] Furthermore, during step S207, the setting unit 42 writes the prior knowledge 34 used in the exploration, the parameters explored, the exploration range, the variation range during the exploration, etc., to the exploration history 35 of the storage unit 30 for storage. Furthermore, the setting unit 42 writes the parameters when the evaluation value is obtained during the exploration, the values, the evaluation value, etc., to the calculation results 36 of the storage unit 30 for storage.

[0140] Next, the setting unit 42 sets the highest evaluation value as the value of the parameter based on the result of the search in step S207 ( S208 ).

[0141] Next, the setting unit 42 outputs the contents of the exploration history 35 stored in the storage unit 30 to the output device 80 via the input / output I / F 20 (S209). Thus, the value of the set parameter is illustrative to the user, who can estimate how much room there is for improving the evaluation value.

[0142] After step S209 , the control unit 40 ends the parameter setting process S200 .

[0143] In addition, when a setting time is specified when executing the parameter setting process S200, the control unit 40 makes adjustments to set the parameter value within the range of the setting time. Specifically, in step S201, the selection unit 41 derives a number of sample images corresponding to the specified setting time. Moreover, in step S203, the setting unit 42 generates a rough evaluation value map and an evaluation value map using the variation range of the parameter value corresponding to the specified setting time. Furthermore, in steps S204 and S206, the extraction unit 43 extracts a number of parameter candidates and expected parameter candidates corresponding to the specified setting time.

[0144] The above is a description of an exemplary embodiment of the present invention. According to an image processing device 50, an image processing program, and an image processing method of an embodiment of the present invention, the value of the parameter is set based on the prior knowledge 34 related to the parameter, a plurality of processing images, an evaluation item corresponding to the plurality of processing images, and a target value of the evaluation item. Thus, for example, for a certain parameter, a value of an appropriate range can be explored with an appropriate change range, so that compared with a case without prior knowledge, the value of the appropriate parameter can be set in a short time. Therefore, the values ​​of multiple parameters can be set efficiently.

[0145] The embodiments described above are intended to facilitate understanding of the present invention and are not intended to limit the present invention. The various elements and their configurations, materials, conditions, shapes, and sizes included in the embodiments are not limited to those illustrated, but can be appropriately changed. Moreover, the structures shown in different embodiments can be partially replaced or combined with each other.

[0146] (Note)

[0147] 1. An image processing device 50 processes an image for processing based on a plurality of parameters, the image processing device 50 comprising:

[0148] a prior knowledge generating unit 44 that generates the prior knowledge 34 related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; and

[0149] The setting unit 42 sets the value of the parameter based on the prior knowledge 34, a plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target values ​​of the evaluation items.

[0150] 8. An image processing program, executed by a computer, for processing an image for processing based on a plurality of parameters, the image processing program comprising:

[0151] a prior knowledge generating step of generating prior knowledge 34 related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; and

[0152] The setting step sets the value of the parameter based on the prior knowledge 34, a plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target values ​​of the evaluation items.

[0153] 15. An image processing method, which processes an image for processing based on multiple parameters, the image processing method comprising:

[0154] a prior knowledge generating step of generating prior knowledge 34 related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; and

[0155] The setting step sets the value of the parameter based on the prior knowledge 34, a plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target values ​​of the evaluation items.

Claims

1. An image processing device for processing an image based on a plurality of parameters, the image processing device comprising: a prior knowledge generating unit that generates prior knowledge related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; as well as The setting unit sets the value of the parameter based on the prior knowledge, the plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target value of the evaluation item.

2. The image processing apparatus according to claim 1, wherein The prior knowledge generation unit generates the prior knowledge based on an evaluation component for obtaining an evaluation value of the evaluation function corresponding to the value of the parameter and a plurality of the evaluation values ​​obtained by using the evaluation component, using an evaluation function determined according to the evaluation item and the target value of the evaluation item. The setting unit sets the value of the parameter so that the evaluation value becomes higher.

3. The image processing apparatus according to claim 2, wherein The evaluation means is an evaluation value map generated using the evaluation function and indicating the evaluation value of the value of each of the parameters in the evaluation item.

4. The image processing device according to any one of claims 1 to 3, wherein The prior knowledge generating unit generates the prior knowledge for each type of the processing or each type of the object of the processing image. The image processing device further includes a selection unit configured to select the prior knowledge based on the designated type or the designated category.

5. The image processing device according to claim 2 or 3, wherein The prior knowledge includes at least one of the degree of influence of the parameter on the evaluation item, the degree of dependency between the parameters, the magnitude of change in the value of the parameter, and the range of change in the value of the parameter.

6. The image processing device according to claim 2 or 3, wherein The setting unit sets, as the parameter, a value having the highest evaluation value within a specified set time range.

7. The image processing device according to any one of claims 1 to 3, further comprising: an extracting unit, extracting at least one of the plurality of parameters based on the prior knowledge, The setting unit sets the value of the extracted parameter.

8. A computer-readable storage medium storing an image processing program, wherein the image processing program is executed by a computer and processes an image for processing based on a plurality of parameters, the image processing program comprising: a prior knowledge generating step of generating prior knowledge related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; as well as The setting step sets the value of the parameter based on the prior knowledge, the plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target value of the evaluation item.

9. The computer-readable storage medium of claim 8, wherein The prior knowledge generation step includes: using an evaluation function determined according to the evaluation item and the target value of the evaluation item, generating the prior knowledge based on an evaluation component for obtaining an evaluation value of the evaluation function corresponding to the value of the parameter, and a plurality of the evaluation values ​​obtained by using the evaluation component, The setting step includes setting the value of the parameter in such a way that the evaluation value becomes higher.

10. The computer-readable storage medium of claim 9, wherein The evaluation means is an evaluation value map generated using the evaluation function and indicating the evaluation value of the value of each of the parameters in the evaluation item.

11. The computer-readable storage medium according to any one of claims 8 to 10, wherein The prior knowledge generation step includes generating the prior knowledge for each type of the processing or each type of the object of the processing image, The image processing program also includes: A selecting step of selecting the prior knowledge based on the designated type or variety.

12. The computer-readable storage medium according to claim 9 or 10, wherein The prior knowledge includes at least one of the degree of influence of the parameter on the evaluation item, the degree of dependency between the parameters, the magnitude of change in the value of the parameter, and the range of change in the value of the parameter.

13. The computer-readable storage medium according to claim 9 or 10, wherein The setting step includes setting the highest evaluation value within a specified setting time range as the parameter.

14. The computer-readable storage medium according to any one of claims 8 to 10, further comprising: an extraction step of extracting at least one of the plurality of parameters based on the prior knowledge, The setting step sets the value of the extracted parameter.

15. An image processing method, performing processing on an image for processing based on a plurality of parameters, the image processing method comprising: a prior knowledge generating step of generating prior knowledge related to the parameter using a plurality of learning images, evaluation items corresponding to the plurality of learning images, and target values ​​of the evaluation items; as well as The setting step sets the value of the parameter based on the prior knowledge, the plurality of processing images, the evaluation items corresponding to the plurality of processing images, and the target value of the evaluation item.

Images