CN111222419A - Object identification method, robot and computer readable storage medium - Google Patents
Object identification method, robot and computer readable storage medium Download PDFInfo
- Publication number
- CN111222419A CN111222419A CN201911350891.2A CN201911350891A CN111222419A CN 111222419 A CN111222419 A CN 111222419A CN 201911350891 A CN201911350891 A CN 201911350891A CN 111222419 A CN111222419 A CN 111222419A
- Authority
- CN
- China
- Prior art keywords
- picture
- confidence
- robot
- recognized
- foreground
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000003860 storage Methods 0.000 title claims abstract description 15
- 238000012545 processing Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 2
- 230000007613 environmental effect Effects 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 22
- 238000010586 diagram Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 239000003086 colorant Substances 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 241000675108 Citrus tangerina Species 0.000 description 2
- 241000220225 Malus Species 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 244000276331 Citrus maxima Species 0.000 description 1
- 235000001759 Citrus maxima Nutrition 0.000 description 1
- 240000000560 Citrus x paradisi Species 0.000 description 1
- 235000021016 apples Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Manipulator (AREA)
Abstract
The invention discloses an object identification method, a robot and a computer readable storage medium, wherein the object identification method is applied to the robot and comprises the following steps: the robot acquires a picture of an object to be identified and judges whether the picture meets a preset condition; if the picture meets the preset conditions, identifying the object to be identified in the picture, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object type; determining and outputting the object type of the object to be recognized in the picture according to the confidence coefficient of the foreground and the confidence coefficient of the object to be recognized relative to each object type; through the mode, the problem of inaccurate recognition result caused by the self-movement of the robot and environmental factors can be solved to a certain extent, and the accuracy of robot character body recognition is improved.
Description
Technical Field
The present invention relates to the field of robot technology, and in particular, to an object recognition method, a robot, and a computer-readable storage medium.
Background
With the development of science and technology, the object recognition technology of pictures is applied more and more widely in the life of people. People only need to shoot and upload the object to the cloud end identification, can obtain the classification information of the object that returns from the cloud end, convenient and fast. Meanwhile, the technology of object recognition of pictures is also gradually applied in the field of robots.
At present, most of robot figure body recognition functions that fall to the ground on the market all set up in the high in the clouds, and the robot acquires the picture of waiting to discern the object through local camera earlier, uploads the picture of waiting to discern the object and reaches the high in the clouds again, and the high in the clouds is received the picture after, discerns the picture again, returns the discernment result at last and gives the robot.
However, the object recognition method provided in the cloud has the following problems: when network delay exists, the experience of the object identification function is poor, and the identification process is slow; when no network exists, the object identification function arranged at the cloud end cannot be used; meanwhile, factors such as the motion of the robot and the environment where the robot is located can increase the inaccuracy rate of the object recognition result, such as: the motion of the robot causes the image shot by the camera to be quite blurred, the object recognized by the robot is in a backlight environment, the ambient light where the robot is located is not white, and the like.
Disclosure of Invention
The invention provides an object recognition method, a robot and a computer readable storage medium, which are used for improving the accuracy of an object recognition result of a robot object recognition function.
In order to solve the above technical problem, the present invention provides an object recognition method applied to a robot, wherein the object recognition method includes: the robot acquires a picture of an object to be identified and judges whether the picture meets a preset condition; if the picture meets the preset conditions, identifying the object to be identified in the picture, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object category; and determining and outputting the object class of the object to be recognized in the picture according to the confidence coefficient of the foreground and the confidence coefficient of the object to be recognized relative to each object class.
If the picture meets the preset conditions, identifying the object to be identified in the picture, and determining the foreground confidence coefficient of the picture and the confidence coefficient of the object to be identified relative to each object category comprises the following steps: and if the picture meets the preset conditions, identifying the object to be identified in the picture by using an offline object identification algorithm, and determining the foreground confidence coefficient of the picture and the confidence coefficient of the object to be identified relative to each object class in the object classes identifiable by the robot.
The step of determining and outputting the object type of the object to be recognized in the picture according to the confidence coefficient of the foreground and the confidence coefficient of the object to be recognized relative to each object type specifically comprises the following steps: if the foreground confidence reaches the foreground threshold, judging whether the maximum confidence in the object class confidence reaches the class threshold; if the object class reaches the class threshold, outputting the object class with the maximum confidence coefficient; if the category threshold is not reached, the specified content is output.
The object identification method further comprises the following steps: and if the foreground confidence coefficient does not reach the foreground threshold value, outputting the specified content.
The robot acquires a picture of an object to be identified, and the step of judging whether the picture meets a preset condition comprises the following steps: judging whether the definition, the chromaticity and the brightness of the picture meet respective corresponding preset conditions; and if the definition, the chromaticity and the brightness of the picture meet the respective corresponding preset conditions, determining that the picture meets the preset conditions.
The object identification method further comprises the following steps: if the brightness of the picture does not meet the corresponding preset condition, judging whether the brightness is within the range of a preset processing threshold value; and if the brightness is within the preset processing threshold range, processing the picture through the high dynamic range image, and then determining that the picture meets the preset condition.
Wherein, still include before the step of judging whether definition, colourity and luminance of picture satisfy respective corresponding preset condition: and converting the picture into an RGB format, cutting the converted picture, and removing the edge background.
Wherein, convert the picture into RGB format, and cut out the picture after will changing, still include after the step of removing the edge background: and adjusting the size of the picture to accord with the preset calculation size of the off-line object recognition algorithm.
In order to solve the above technical problem, the present invention further provides a robot, including: the device comprises a processor and a memory, wherein the memory stores pictures of objects to be identified and identifiable object classes, and the processor is used for executing the object identification method.
In order to solve the above technical problem, the present invention also provides a computer-readable storage medium storing program data, which can be executed to implement the above object recognition method.
The invention has the beneficial effects that: different from the situation of the prior art, the method firstly judges whether the picture of the object to be recognized meets the preset condition, recognizes the picture after the picture meets the preset condition, determines the foreground confidence coefficient of the picture and the confidence coefficient of the object to be recognized relative to each object type, determines the object type corresponding to the maximum confidence coefficient as the type of the object to be recognized, wherein the maximum confidence coefficient in the foreground confidence coefficient and the confidence coefficient of the object type is larger than the corresponding preset threshold value. Through the mode, the problem of inaccurate recognition result caused by the self-movement of the robot and environmental factors can be solved to a certain extent, and the accuracy of robot character body recognition is improved.
Drawings
FIG. 1 is a schematic flow chart diagram illustrating an embodiment of an object recognition method according to the present invention;
FIG. 2 is a schematic flow chart of another embodiment of an object recognition method provided by the present invention;
FIG. 3 is a schematic structural diagram of an embodiment of a robot provided by the present invention;
fig. 4 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided in the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, fig. 1 is a schematic flow chart of an embodiment of an object identification method according to the present invention. The object recognition method of the present invention is applied to a robot.
S11: the robot acquires a picture of the object to be identified and judges whether the picture meets a preset condition.
The object recognition function of the present embodiment is provided in a robot, and is an off-line object recognition technique. When the main program of the robot detects that the object recognition function of the robot is started, the robot shoots a picture of an object to be recognized through a robot camera and judges whether the quality of the picture of the object to be recognized meets the preset condition of the object recognition function of the robot.
S12: and if the picture meets the preset conditions, identifying the object to be identified in the picture, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object type.
And if the robot judges that the picture of the object to be recognized accords with the preset condition of the robot character recognition function, the robot recognizes the object to be recognized in the picture, and determines the foreground confidence coefficient of the picture and the confidence coefficient of the object to be recognized relative to each object category. Wherein the confidence of each object class refers to the confidence of the object to be recognized relative to each object class in all object classes which can be recognized by the robot, and the confidence of each object class is added up to be equal to 1. The foreground means that the classification of the object to be recognized in the picture belongs to a target object, if the object is not the target object, the classification of the object to be recognized in the picture is a background, that is, the object is not the object to be recognized, and the classification of the object to be recognized here includes: foreground and background, where the foreground confidence is added to the background confidence also equal to 1.
Wherein, in statistics, the Confidence interval (Confidence interval) of a probability sample is an interval estimate of some overall parameter of the sample. The confidence interval exhibits the extent to which the true value of this parameter has a certain probability of falling around the measurement. The confidence interval indicates the degree of plausibility of the measured value of the measured parameter, i.e. the "certain probability" required above. This probability is called the confidence.
S13: and determining and outputting the object class of the object to be recognized in the picture according to the confidence coefficient of the foreground and the confidence coefficient of the object to be recognized relative to each object class.
The specific steps of determining and outputting the object type of the object to be recognized in the picture according to the confidence coefficient of the foreground and the confidence coefficient of the object to be recognized relative to each object type are as follows: and judging whether the foreground confidence obtained in the last step reaches a foreground threshold, if the foreground confidence reaches the foreground threshold, determining that the foreground exists in the picture, namely the target object of the object to be recognized, and further judging whether the maximum confidence in the confidence of the object class obtained in the last step reaches the class threshold. If the confidence of the foreground does not reach the foreground threshold, the foreground does not exist in the picture, namely the picture of the object to be recognized is shot wrongly, the robot outputs specified contents, the specified contents comprise other contents or the contents such as the shot mistakes, and the setting can be carried out according to the practical application. The foreground threshold and the category threshold can be set according to actual application so as to meet the quality requirement of the actual application.
And the maximum confidence coefficient in the confidence coefficients of the object classes refers to the confidence coefficient of the object class with the maximum value in the confidence coefficients of the object to be recognized, which is obtained in the last step, relative to each object class in all the object classes which can be recognized by the robot, and the maximum confidence coefficient is judged to judge whether the maximum confidence coefficient reaches a class threshold value or not.
And if the maximum confidence reaches the category threshold, determining the object category corresponding to the maximum confidence as the object category of the object to be recognized, and outputting the object category with the maximum confidence. If the maximum confidence does not reach the category threshold, the robot can recognize the object categories without the objects to be recognized in the object categories, and specified contents, such as other categories and the like, are output. Wherein, the category threshold value can be set according to the practical application condition.
Through the mode, the object recognition function is arranged on the robot, the image preprocessing step and the step of determining and judging the foreground confidence of the image and the confidence of each object type are added to recognize the object to be recognized, so that the problem that the robot cannot normally use the object recognition function due to network reasons is solved, the image of the object to be recognized is effectively screened, the quality of the recognized image is improved, the problem that the object recognition result is inaccurate due to the self motion of the robot and environmental factors is solved to a certain extent, and the recognition accuracy of the object recognition function of the robot is improved.
Referring to fig. 2, fig. 2 is a schematic flow chart of an object identification method according to another embodiment of the present invention. The object recognition method of the present invention is applied to a robot.
S21: the robot acquires a picture of an object to be recognized, converts the picture into an RGB format, cuts the converted picture and removes an edge background.
The object recognition function of the present embodiment is provided in a robot, and is an off-line object recognition technique. When a main program of the robot detects that an object recognition function of the robot is started, the robot shoots a picture of an object to be recognized through a robot camera, converts the color format of the picture of the object to be recognized into an RGB format, and then cuts the picture of the object to be recognized converted into the RGB format according to a certain proportion so as to remove irrelevant backgrounds at the edge and retain the object to be recognized at the central position, thereby improving the recognition accuracy. The RGB color scheme is a color standard in the industry, and various colors are obtained by changing three color channels of red (R), green (G), and blue (B) and superimposing them on each other, where RGB represents the colors of the three channels of red, green, and blue, and the standard includes almost all colors that can be perceived by human vision.
S22: and adjusting the size of the picture to accord with the preset calculation size of the off-line object recognition algorithm.
And adjusting the size of the cut picture of the object to be recognized to enable the picture to conform to the preset size of the off-line object recognition algorithm, wherein the preset size is the size which can be supported by the off-line object recognition algorithm, so that the subsequent off-line object recognition algorithm can calculate the picture of the object to be recognized.
S23: judging whether the definition, the chromaticity and the brightness of the picture meet respective corresponding preset conditions; and if the definition, the chromaticity and the brightness of the picture meet the respective corresponding preset conditions, determining that the picture meets the preset conditions.
Judging whether the definition of the picture of the object to be identified reaches a definition threshold value or not, and if the definition of the picture does not reach the definition threshold value, returning to the picture abnormity; if the definition of the picture reaches a definition threshold value, further judging whether the chroma of the picture of the object to be identified is within a preset chroma range, and if the chroma of the picture of the object to be identified is not within the preset chroma range, returning to the picture to be abnormal; if the chroma of the picture of the object to be recognized is in the preset chroma range, further judging whether the chroma of the picture of the object to be recognized is in the preset brightness range, and if the chroma of the picture of the object to be recognized is in the preset brightness range, determining that the picture meets the preset condition; if the brightness of the picture of the object to be recognized is not within the preset brightness range, further judging whether the brightness of the picture of the object to be recognized is within the preset processing threshold range, and if the brightness of the picture of the object to be recognized is not within the preset processing threshold range, returning to the picture to be abnormal; and if the brightness of the picture of the object to be recognized is within the preset processing threshold range, processing the picture through the high dynamic range image, and then determining that the picture meets the preset condition. And after the picture abnormity is returned in the process, the result of the picture abnormity is returned to the main program of the robot. The preset processing threshold range refers to a processable threshold range in which the robot can process the picture in a high dynamic range image mode to enable the picture to meet a preset condition.
Compared with common images, High-Dynamic Range (HDR) images can provide more Dynamic ranges and image details, and according to LDR (Low-Dynamic Range) images with different exposure times, LDR images with optimal details corresponding to each exposure time are used to synthesize and synthesize a final HDR image, so that visual effects of people in a real environment can be better reflected.
S24: and identifying the object to be identified in the picture by using an offline object identification algorithm, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object class in the robot identifiable object classes.
And identifying the picture of the object to be identified which meets the preset condition by using an offline object identification algorithm, and determining the foreground confidence coefficient of the picture and the confidence coefficient of the object to be identified relative to each object category. Wherein the confidence of each object class refers to the confidence of the object to be recognized relative to each object class in all object classes which can be recognized by the robot, and the confidence of each object class is added up to be equal to 1. The foreground means that the classification of the object to be recognized in the picture belongs to a target object, if the object is not the target object, the classification of the object to be recognized in the picture is a background, that is, the object is not the object to be recognized, and the classification of the object to be recognized here includes: foreground and background, where the foreground confidence is added to the background confidence also equal to 1. Wherein, in statistics, the Confidence interval (Confidence interval) of a probability sample is an interval estimate of some overall parameter of the sample. The confidence interval exhibits the extent to which the true value of this parameter has a certain probability of falling around the measurement. The confidence interval indicates the degree of plausibility of the measured value of the measured parameter, i.e. the "certain probability" required above. This probability is called the confidence.
Specifically, the confidence of the object to be recognized with respect to each of all object classes that the robot can recognize is exemplified by: assume that the robot can recognize 5 object categories: apple, tangerine, shaddock, ball and cup, then supposing that the object to be discerned is the tangerine, then the confidence coefficient that the object to be discerned can discern every kind of object classification in all object classifications for the robot respectively is: 3% of apples, 93% of oranges, 4% of grapefruits, 0% of ball and 0% of cups, and the confidence coefficient of the object types which can be recognized by the robot is calculated. I.e., the sum of the five sets of confidence levels is 1. And each object class refers to each of the object classes that the robot can recognize. In practical applications, the type of the object that can be recognized by the robot is not limited to this.
S25: and if the foreground confidence coefficient reaches the foreground threshold value, judging whether the maximum confidence coefficient in the confidence coefficients of the object classes reaches the class threshold value.
And judging whether the foreground confidence coefficient obtained in the last step reaches a foreground threshold value, if the foreground confidence coefficient reaches the foreground threshold value, determining that the foreground exists in the picture, namely the target object of the object to be recognized, and further judging whether the maximum confidence coefficient in the confidence coefficients of the object classes obtained in the last step reaches the class threshold value. If the confidence of the foreground does not reach the foreground threshold, the foreground does not exist in the picture, namely the picture of the object to be recognized is shot wrongly, the robot outputs specified contents, the specified contents comprise other contents or the contents such as the shot mistakes, and the setting can be carried out according to the practical application. The foreground threshold and the category threshold can be set according to actual application so as to meet the quality requirement of the actual application.
And the maximum confidence coefficient in the confidence coefficients of the object classes refers to the confidence coefficient of the object class with the maximum value in the confidence coefficients of the object to be recognized, which is obtained in the last step, relative to each object class in all the object classes which can be recognized by the robot, and the maximum confidence coefficient is judged to judge whether the maximum confidence coefficient reaches a class threshold value or not. Further, the example in the above step is explained, and if the category threshold is set to 80%, in the example in the previous step, the category threshold is set to 80% when the maximum confidence level orange 93% is reached.
S26: if the object class reaches the class threshold, outputting the object class with the maximum confidence coefficient; if the category threshold is not reached, the specified content is output.
And if the maximum confidence reaches the category threshold, determining the object category corresponding to the maximum confidence as the object category of the object to be recognized, and outputting the object category with the maximum confidence. In the above example, the object type orange corresponding to 93% of the maximum confidence orange is determined as the type of the object to be recognized, and the type orange is output. If the maximum confidence does not reach the category threshold, the robot can recognize the object categories without the objects to be recognized in the object categories, and specified contents, such as other categories and the like, are output. In this step, the recognition result is returned to the main program of the robot regardless of the recognition result.
Through the mode, the object recognition function is arranged on the robot, the image preprocessing step and the step of determining and judging the foreground confidence of the image and the confidence of each object type are added to recognize the object to be recognized, so that the problem that the robot cannot normally use the object recognition function due to network reasons is solved, the image of the object to be recognized is effectively screened, the quality of the recognized image is improved, the problem that the object recognition result is inaccurate due to the self motion of the robot and environmental factors is solved to a certain extent, and the recognition accuracy of the object recognition function of the robot is improved.
Based on the same inventive concept, the present invention further provides a robot, which can be executed to implement the object identification method according to any of the above embodiments, please refer to fig. 3, fig. 3 is a schematic structural diagram of an embodiment of the robot provided by the present invention, and the robot includes a processor 41 and a memory 42.
The memory 42 is used for storing a picture of an object to be identified, a definition threshold, a preset chromaticity range, a preset brightness range, a processable range, a foreground threshold, a category threshold, and a recognizable object category feature library.
The processor 41 is configured to, when detecting that the object recognition function of the robot is activated, take a picture of an object to be recognized through the robot camera, and determine whether the quality of the picture of the object to be recognized meets a preset condition of the object recognition function of the robot.
If the processor 41 judges that the picture of the object to be recognized meets the preset condition of the robot body recognition function, the processor 41 recognizes the object to be recognized in the picture, and determines the foreground confidence of the picture and the confidence of the object to be recognized relative to each object type. Wherein the confidence of each object class refers to the confidence of the object to be recognized relative to each object class in all object classes which can be recognized by the robot, and the confidence of each object class is added up to be equal to 1. The foreground means that the classification of the object to be recognized in the picture belongs to a target object, if the object is not the target object, the classification of the object to be recognized in the picture is a background, that is, the object is not the object to be recognized, and the classification of the object to be recognized here includes: foreground and background, where the foreground confidence is added to the background confidence also equal to 1.
The processor 41 further determines whether the foreground confidence reaches a foreground threshold, and if the foreground confidence reaches the foreground threshold, it is determined that the foreground exists in the picture, that is, the object to be recognized, and further determines whether the maximum confidence among the confidence of the object class reaches a class threshold. If the confidence of the foreground does not reach the foreground threshold, the foreground does not exist in the picture, namely the picture of the object to be recognized is shot wrongly, the robot outputs specified contents, the specified contents comprise other contents or the contents such as the shot mistakes, and the setting can be carried out according to the practical application. The foreground threshold and the category threshold can be set according to actual application so as to meet the quality requirement of the actual application.
And the maximum confidence coefficient in the confidence coefficients of the object classes refers to the confidence coefficient of the object class with the maximum value in the confidence coefficients of the object to be recognized, which is obtained in the last step, relative to each object class in all the object classes which can be recognized by the robot, and the maximum confidence coefficient is judged to judge whether the maximum confidence coefficient reaches a class threshold value or not.
And if the maximum confidence reaches the category threshold, determining the object category corresponding to the maximum confidence as the object category of the object to be recognized, and outputting the object category with the maximum confidence. If the maximum confidence does not reach the category threshold, the object category without the object to be recognized in the recognizable object categories or the shot picture has errors, and specified contents such as other categories or shot errors are output.
Through the mode, the object recognition function is arranged on the robot, the image preprocessing step and the step of determining and judging the foreground confidence of the image and the confidence of each object type are added to recognize the object to be recognized, so that the problem that the robot cannot normally use the object recognition function due to network reasons is solved, the image of the object to be recognized is effectively screened, the quality of the recognized image is improved, the problem that the object recognition result is inaccurate due to the self motion of the robot and environmental factors is solved to a certain extent, and the recognition accuracy of the object recognition function of the robot is improved.
Based on the same inventive concept, the present invention further provides a computer-readable storage medium, please refer to fig. 4, and fig. 4 is a schematic structural diagram of an embodiment of the computer-readable storage medium provided in the present invention. The computer-readable storage medium 50 has stored therein at least one program data 51, the program data 51 being adapted to implement any of the methods described above. In one embodiment, the computer-readable storage medium 50 includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
On one hand, the invention separates the instruction cache and the data cache physically, and the physical separation structure improves the data acquisition efficiency and reduces the manufacturing complexity of the microprocessor. On the other hand, the instruction cache and the data cache are logically combined, which is equivalent to enlarging the capacity of the first-level cache, thereby improving the cache hit rate.
In the embodiments provided in the present invention, it should be understood that the disclosed method and apparatus can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a module or a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium, or in a part of or all of the technical solution that contributes to the prior art.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. An object recognition method applied to a robot, the object recognition method comprising:
the robot acquires a picture of an object to be identified and judges whether the picture meets a preset condition;
if the picture meets the preset conditions, identifying the object to be identified in the picture, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object category;
and determining and outputting the object class of the object to be recognized in the picture according to the foreground confidence coefficient and the confidence coefficient of the object to be recognized relative to each object class.
2. The object recognition method according to claim 1, wherein the step of recognizing the object to be recognized in the picture if the picture meets a preset condition, and determining the confidence level of the foreground of the picture and the confidence level of the object to be recognized relative to each object class comprises:
and if the picture meets the preset conditions, identifying the object to be identified in the picture by using an offline object identification algorithm, and determining the foreground confidence of the picture and the confidence of the object to be identified relative to each object class in the robot identifiable object classes.
3. The object recognition method according to claim 1, wherein the step of determining and outputting the object class of the object to be recognized in the picture according to the confidence of the foreground and the confidence of the object to be recognized with respect to each object class specifically comprises:
if the foreground confidence reaches a foreground threshold, judging whether the maximum confidence in the object class confidence reaches a class threshold;
if the category threshold is reached, outputting the object category with the maximum confidence coefficient; and if the category threshold is not reached, outputting the specified content.
4. The object recognition method according to claim 3, characterized in that the object recognition method further comprises:
and if the foreground confidence coefficient does not reach the foreground threshold value, outputting the specified content.
5. The object identification method according to claim 1, wherein the step of acquiring the picture of the object to be identified by the robot and judging whether the picture meets the preset condition comprises the steps of:
judging whether the definition, the chromaticity and the brightness of the picture meet respective corresponding preset conditions;
and if the definition, the chromaticity and the brightness of the picture meet the respective corresponding preset conditions, determining that the picture meets the preset conditions.
6. The object recognition method according to claim 5, characterized in that the object recognition method further comprises:
if the brightness of the picture does not meet the corresponding preset condition, judging whether the brightness is within a preset processing threshold range;
and if the brightness is within the preset processing threshold range, processing the picture through a high dynamic range image, and then determining that the picture meets the preset condition.
7. The object recognition method according to claim 5 or 6, wherein the step of determining whether the definition, the chromaticity and the brightness of the picture satisfy respective preset conditions further comprises:
and converting the picture into an RGB format, cutting the converted picture, and removing the edge background.
8. The object recognition method of claim 7, wherein the steps of converting the picture into RGB format and cropping the converted picture and removing the edge background further comprise:
and adjusting the size of the picture to accord with the preset calculation size of the off-line object recognition algorithm.
9. A robot, characterized in that the robot comprises: a processor and a memory, the memory having stored therein a picture of an object to be identified and an identifiable class of objects, the processor being configured to perform the object identification method of any one of claims 1-8.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores program data executable to implement the object identification method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911350891.2A CN111222419A (en) | 2019-12-24 | 2019-12-24 | Object identification method, robot and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911350891.2A CN111222419A (en) | 2019-12-24 | 2019-12-24 | Object identification method, robot and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111222419A true CN111222419A (en) | 2020-06-02 |
Family
ID=70825965
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911350891.2A Pending CN111222419A (en) | 2019-12-24 | 2019-12-24 | Object identification method, robot and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111222419A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112949423A (en) * | 2021-02-07 | 2021-06-11 | 深圳市优必选科技股份有限公司 | Object recognition method, object recognition device, and robot |
| WO2022082999A1 (en) * | 2020-10-21 | 2022-04-28 | 深圳市优必选科技股份有限公司 | Object recognition method and apparatus, and terminal device and storage medium |
| CN114979473A (en) * | 2022-05-16 | 2022-08-30 | 遥相科技发展(北京)有限公司 | Industrial robot control method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106874845A (en) * | 2016-12-30 | 2017-06-20 | 东软集团股份有限公司 | The method and apparatus of image recognition |
| CN106920229A (en) * | 2017-01-22 | 2017-07-04 | 北京奇艺世纪科技有限公司 | Image obscuring area automatic testing method and system |
| CN108205736A (en) * | 2016-12-16 | 2018-06-26 | 杭州海康威视系统技术有限公司 | A kind of data processing method and device |
| CN108960290A (en) * | 2018-06-08 | 2018-12-07 | Oppo广东移动通信有限公司 | Image processing method, device, computer readable storage medium and electronic equipment |
| CN109858436A (en) * | 2019-01-29 | 2019-06-07 | 中国科学院自动化研究所 | Target category modification method, detection method based on video dynamic foreground mask |
| CN110321935A (en) * | 2019-06-13 | 2019-10-11 | 上海上湖信息技术有限公司 | Business event relationship determines method and device, computer readable storage medium |
-
2019
- 2019-12-24 CN CN201911350891.2A patent/CN111222419A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108205736A (en) * | 2016-12-16 | 2018-06-26 | 杭州海康威视系统技术有限公司 | A kind of data processing method and device |
| CN106874845A (en) * | 2016-12-30 | 2017-06-20 | 东软集团股份有限公司 | The method and apparatus of image recognition |
| CN106920229A (en) * | 2017-01-22 | 2017-07-04 | 北京奇艺世纪科技有限公司 | Image obscuring area automatic testing method and system |
| CN108960290A (en) * | 2018-06-08 | 2018-12-07 | Oppo广东移动通信有限公司 | Image processing method, device, computer readable storage medium and electronic equipment |
| WO2019233266A1 (en) * | 2018-06-08 | 2019-12-12 | Oppo广东移动通信有限公司 | Image processing method, computer readable storage medium and electronic device |
| CN109858436A (en) * | 2019-01-29 | 2019-06-07 | 中国科学院自动化研究所 | Target category modification method, detection method based on video dynamic foreground mask |
| CN110321935A (en) * | 2019-06-13 | 2019-10-11 | 上海上湖信息技术有限公司 | Business event relationship determines method and device, computer readable storage medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022082999A1 (en) * | 2020-10-21 | 2022-04-28 | 深圳市优必选科技股份有限公司 | Object recognition method and apparatus, and terminal device and storage medium |
| CN112949423A (en) * | 2021-02-07 | 2021-06-11 | 深圳市优必选科技股份有限公司 | Object recognition method, object recognition device, and robot |
| CN112949423B (en) * | 2021-02-07 | 2024-05-24 | 深圳市优必选科技股份有限公司 | Object recognition method, object recognition device and robot |
| CN114979473A (en) * | 2022-05-16 | 2022-08-30 | 遥相科技发展(北京)有限公司 | Industrial robot control method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR100568237B1 (en) | Apparatus and method for extracting moving objects from video image | |
| CN111222419A (en) | Object identification method, robot and computer readable storage medium | |
| CN109361910B (en) | Self-adaptive white balance correction method and device | |
| CN107680056B (en) | Image processing method and device | |
| US20030021478A1 (en) | Image processing technology for identification of red eyes in image | |
| US10949984B2 (en) | Object tracking system and method thereof | |
| CN112348778B (en) | Object identification method, device, terminal equipment and storage medium | |
| CN107808109B (en) | Two-dimensional code image identification method and mobile terminal | |
| US8817125B2 (en) | Gesture recognition using chroma-keying | |
| CN111028276A (en) | Image alignment method and device, storage medium and electronic equipment | |
| WO2020119454A1 (en) | Method and apparatus for color reproduction of image | |
| CN111800294B (en) | Gateway fault diagnosis method and device, network equipment and storage medium | |
| CN112052726A (en) | Image processing method and device | |
| CN111970501A (en) | Pure color scene AE color processing method and device, electronic equipment and storage medium | |
| CN109598195B (en) | Method and device for processing clear face image based on monitoring video | |
| CN115334250B (en) | Image processing method and device and electronic equipment | |
| CN112949423B (en) | Object recognition method, object recognition device and robot | |
| CN114037741B (en) | Self-adaptive target detection method and device based on event camera | |
| CN116668843A (en) | Shooting state switching method and device, electronic equipment and storage medium | |
| CN113613024B (en) | Video preprocessing method and device | |
| CN111859022A (en) | Cover generation method, electronic device and computer-readable storage medium | |
| CN105141857A (en) | Image processing method and device | |
| JP7412947B2 (en) | Image processing device, image processing method and program | |
| CN114219744B (en) | Image generation method, device, equipment and storage medium | |
| CN109389582B (en) | Method and device for identifying brightness of image main body |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |