WO2023045185A1 - 对象检测方法及装置、电子设备和存储介质 - Google Patents

对象检测方法及装置、电子设备和存储介质 Download PDF

Info

Publication number
WO2023045185A1
WO2023045185A1 PCT/CN2022/071050 CN2022071050W WO2023045185A1 WO 2023045185 A1 WO2023045185 A1 WO 2023045185A1 CN 2022071050 W CN2022071050 W CN 2022071050W WO 2023045185 A1 WO2023045185 A1 WO 2023045185A1
Authority
WO
WIPO (PCT)
Prior art keywords
detection
image
area
attribute
object detection
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.)
Ceased
Application number
PCT/CN2022/071050
Other languages
English (en)
French (fr)
Inventor
杨昆霖
熊思龙
侯军
伊帅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Sensetime Intelligent Technology Co Ltd
Original Assignee
Shanghai Sensetime Intelligent Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Sensetime Intelligent Technology Co Ltd filed Critical Shanghai Sensetime Intelligent Technology Co Ltd
Publication of WO2023045185A1 publication Critical patent/WO2023045185A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Definitions

  • the present disclosure relates to the field of computer technology, and in particular to an object detection method and device, electronic equipment, and a storage medium.
  • the disclosure proposes an object detection method and device, electronic equipment, and a storage medium, which can timely and effectively correct the detection result without correcting the algorithm, and improve the accuracy of the detection result.
  • an object detection method comprising:
  • the false detection region attribute being used to characterize the location and size of the false detection region
  • At least one object detection result includes detection area attributes representing the position and size of the detection object;
  • the object detection result is screened according to the false detection area attribute and the detection area attribute to obtain an intrusion detection result.
  • the determining at least one false detection region attribute includes:
  • At least one attribute of the false detection region is determined.
  • the false detection region generated in response to the human-computer interaction operation is implemented in the following manner:
  • the pre-detection result including at least one preliminary detection frame
  • the false detection region generated in response to the human-computer interaction operation is implemented in the following manner:
  • determining that the selected area is a falsely detected area includes:
  • the determining at least one image frame corresponding to the image to be edited includes:
  • the object detection result includes a detection frame coordinate and a corresponding probability value, and the probability value is used to represent the possibility that the target object exists in the image to be detected within the corresponding detection frame coordinate,
  • the detection frame coordinates are the attributes of the detection area characterized by coordinates.
  • the filtering the object detection result according to the false detection area attribute and the detection area attribute, and obtaining the intrusion detection result includes:
  • the intrusion detection result is determined according to the object detection results obtained after the first screening and the second screening.
  • the first screening of the object detection result according to the corresponding relationship between the attribute of the false detection area and the attribute of the detection area includes;
  • the detection area attribute corresponding to the object in the object detection result is deleted.
  • the responding to the fact that the object in the object detection result is located inside the false detection region represented by the false detection region attribute is implemented in the following manner:
  • performing the second screening according to the probability value in the object detection result includes:
  • the object detection result is deleted.
  • the method further includes:
  • an object detection device comprising:
  • An attribute determination module configured to determine at least one attribute of a false detection region, where the attribute of the false detection region is used to characterize the position and size of the false detection region;
  • An image detection module configured to perform object detection on the image to be detected to obtain at least one object detection result, wherein the at least one object detection result includes detection area attributes representing the position and size of the detection object;
  • the result screening module is configured to screen the object detection result according to the false detection area attribute and the detection area attribute to obtain an intrusion detection result.
  • the attribute determination module includes:
  • the human-computer interaction sub-module is configured to determine at least one false detection area attribute in response to the false detection area generated by the human-computer interaction operation.
  • the human-computer interaction submodule includes:
  • a pre-detection unit configured to perform object detection on the image to be detected and display a pre-detection result of the object detection, the pre-detection result including at least one preliminary detection frame;
  • the first area determining unit is configured to, in response to receiving at least one preliminary detection frame selected by the user through human-computer interaction, determine the area represented by the preliminary detection frame as a false detection area.
  • the human-computer interaction submodule includes:
  • the edit control display unit is used to display the edit control on the scene display interface
  • An editing image display unit configured to display an image to be edited corresponding to the image to be detected in response to the editing control being triggered;
  • the second area determination unit is configured to, in response to at least one area in the image to be edited being selected, determine that the selected area is a falsely detected area.
  • the second area determination unit includes:
  • an image frame determining subunit configured to determine at least one image frame corresponding to the image to be edited
  • the false detection area determination subunit is configured to determine the area within the at least one image frame as the selected false detection area.
  • the image frame determining subunit includes:
  • a coordinate receiving subunit configured to, in response to receiving at least one image frame coordinate, determine an image frame corresponding to the at least one image frame coordinate on the image to be edited;
  • the gesture determining subunit is configured to, in response to receiving at least one preset gesture corresponding to the image to be edited, project at least one region determined on the image to be edited according to the at least one preset gesture as an image frame.
  • the object detection result includes a detection frame coordinate and a corresponding probability value, and the probability value is used to represent the possibility that the target object exists in the image to be detected within the corresponding detection frame coordinate,
  • the detection frame coordinates are the attributes of the detection area characterized by coordinates.
  • the result screening module includes:
  • the first screening submodule is configured to perform a first screening on the object detection result according to the corresponding relationship between the attribute of the false detection area and the attribute of the detection area;
  • the second screening submodule is used to perform a second screening on the object detection result according to the probability value in the object detection result;
  • the result determination submodule is configured to determine the intrusion detection result according to the object detection results obtained after the first screening and the second screening.
  • the first screening submodule includes;
  • the position screening unit is configured to delete the detection area attribute corresponding to the object in the object detection result in response to the object in the object detection result being located inside the false detection area represented by the false detection area attribute.
  • the location screening unit includes:
  • a key point determining subunit configured to determine the key point position of the object in the object detection result
  • the positional relationship determination subunit is configured to determine that the detection area represented by the attribute of the detection area is inside the false detection area in response to the position of the key point being inside a false detection area represented by a false detection area attribute.
  • the second screening submodule includes:
  • a threshold screening unit configured to delete the object detection result in response to a probability value in the object detection result being less than a preset threshold.
  • the device further includes:
  • An image tagging module configured to tag the image to be detected by the intrusion detection result
  • the information display module is used to output the labeled image to be detected and corresponding prompt information.
  • an electronic device including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to call the instructions stored in the memory to execute the above method.
  • a computer-readable storage medium on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the above method is implemented.
  • a computer program product including computer readable codes, or a non-volatile computer readable storage medium bearing computer readable codes, when the computer readable codes are stored in an electronic device
  • the processor in the electronic device is used to implement the above method.
  • the object detection method defines a false detection area in a specific scene, and filters the detection results obtained based on the false detection area during object detection to improve the accuracy of the detection result. Furthermore, in scenarios such as product demonstrations, this method can also be used to detect and identify false positive results and correct errors in time.
  • FIG. 1 shows a flowchart of an object detection method according to an embodiment of the present disclosure
  • Fig. 2 shows a flow chart of the process of determining the attribute of a false detection area according to an embodiment of the present disclosure
  • Fig. 3 shows a schematic diagram of a scene display interface according to an embodiment of the present disclosure
  • Fig. 4 shows a schematic diagram of determining an object detection result according to an embodiment of the present disclosure
  • Fig. 5 shows a schematic diagram of a detection result of a screening object according to an embodiment of the present disclosure
  • Fig. 6 shows a schematic diagram of judging the correspondence between false detection area attributes and detection area attributes according to an embodiment of the present disclosure
  • Fig. 7 shows a schematic diagram of an object detection device according to an embodiment of the present disclosure
  • Fig. 8 is a block diagram of an electronic device according to an exemplary embodiment
  • Fig. 9 is a block diagram of another electronic device according to an exemplary embodiment.
  • Fig. 1 shows a flowchart of an object detection method according to an embodiment of the present disclosure.
  • the object detection method may be executed by electronic equipment such as a terminal device or a server, and the terminal device may be user equipment (User Equipment, UE), mobile device, user terminal, terminal, cellular phone, cordless phone , personal digital assistant (Personal Digital Assistant, PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc.
  • the subject of execution of the steps of the method can be executed by hardware, or executed by a processor running computer-executable code .
  • it may be implemented by a processor calling computer-readable instructions stored in a memory.
  • the method may be performed by a server.
  • the server may be a single server, or a server cluster composed of multiple servers.
  • a terminal device or other processing device can be used to execute the object detection method of the embodiment of the present disclosure to correct the error in the object detection process, thereby improving the object detection process. the accuracy of the results.
  • the object detection method of the embodiment of the present disclosure includes the following steps:
  • Step S10 determining at least one false detection area attribute.
  • the attribute of the false detection area is used to characterize the position and size of the false detection area, for example, may include coordinates of an image frame used to define the false detection area, and the area within the image frame is the false detection area.
  • the false detection area is an area where it is more likely that an object will be falsely detected when there is no object during the detection process.
  • the false detection area may be determined based on historical object detection results or preliminary detection results of the image to be detected.
  • the object detection method in the embodiment of the present disclosure is used for object detection in a fixed scene. That is to say, in the case that false detections often occur in a specific area in the same scene, this area is determined as a false detection area, so that in the subsequent object detection process, the detection results can be filtered through the false detection area attribute of the false detection area, and in Improve the accuracy of detection results without modifying the underlying algorithm.
  • the object detection method in the embodiment of the present disclosure is used to correct the detection result of a specific image to be detected. That is to say, before performing formal object detection on the image to be detected, a preliminary object detection is performed first. In the case of a false detection in the result of the preliminary object detection, the false detection area is determined as a false detection area, so that the detection result can be screened by the false detection area attribute of the false detection area in the formal object detection process, and the formal detection can be improved. the accuracy of the results.
  • the manner of determining the attribute of the false detection area may be to determine at least one attribute of the false detection area in response to the false detection area generated by the human-computer interaction operation. That is to say, the user can directly interact with the electronic device through the human-computer interaction module of the mouse, keyboard, touch screen and other electronic devices, so as to select at least one false detection according to the historical object detection or the preliminary detection results of the image to be detected.
  • the region gets the corresponding false detection region attribute.
  • the process of generating a false detection area in response to human-computer interaction in this embodiment of the present disclosure may further include: performing object detection on the image to be detected and displaying a pre-detection result of the object detection, the pre-detection result including at least A preliminary detection box.
  • determine the region represented by the preliminary detection frame as a false detection region In response to receiving at least one preliminary detection frame selected by the user through human-computer interaction, determine the region represented by the preliminary detection frame as a false detection region. That is to say, before performing formal object detection on the image to be detected, a preliminary object detection may be performed to obtain a pre-detection result, and the pre-detection result includes at least one preliminary detection frame.
  • the electronic device displays the image to be detected with the preliminary detection frame through the display device, and the user can directly identify whether there is an object in each preliminary detection frame, and directly select the preliminary detection frame if it does not exist, so that the area in the preliminary detection frame as the false detection area.
  • Fig. 2 shows a flowchart of a process of determining attributes of false detection regions according to an embodiment of the present disclosure. As shown in FIG. 2, the process of determining the attribute of the false detection area in the embodiment of the present disclosure through human-computer interaction includes the following steps:
  • Step S11 displaying an editing control on the scene display interface.
  • the scene display interface is used to display editing controls.
  • the edit control is used to enter the edit mode when triggered, so as to select the false detection area by editing.
  • the scene display interface is also used to display the image to be edited corresponding to the image to be detected, and the image information corresponding to the image to be edited.
  • the image to be edited corresponds to the image to be detected in this object detection process, and the user can select a false detection area based on the image to be edited after triggering the editing control.
  • the scene display interface may also display content such as image modification controls for replacing, adding, and deleting images to be edited.
  • Step S12 in response to the editing control being triggered, displaying an image to be edited corresponding to the image to be detected.
  • the user may click on the touch screen of the electronic device through a preset gesture, and trigger the editing control through a mouse click or double click.
  • the editing control When the editing control is triggered, the editing mode is entered, that is, the image to be edited is displayed on the scene display interface.
  • the image to be edited is not redisplayed, and the image to be edited is directly converted from a non-editable state to an editable state.
  • the image to be edited is not displayed in the scene display interface, the image to be edited corresponding to the image to be detected is loaded and displayed, and the state of the image to be edited is set as an editable state.
  • the image to be edited may be the same image as the scene where the image to be detected needs to be detected, and may be selected according to different application scenarios. For example, in an application scenario where formal object detection is performed on an image that has undergone preliminary object detection, the image to be detected may be directly determined as the image to be edited. In the application scenario where the object detection result of the image to be detected is corrected according to the detection result of the historical image, the image collected in the same acquisition scene and for which the object detection has been completed can be obtained as the image to be edited.
  • the collection scene may be scenes such as stairwells, streets, corridors, and residential areas.
  • the image to be edited also displays the location of the object detected after historical object detection or preliminary object detection, and the location can be marked by a corresponding detection frame.
  • Step S13 in response to at least one area in the image to be edited being selected, determining that the selected area is a falsely detected area.
  • the incorrectly detected area is determined by selecting a partial area in the image to be edited.
  • a partial area in the image to be edited is selected by adding an image frame on the image to be edited. That is to say, at least one image frame corresponding to the image to be edited is determined, and an area within the at least one image frame is determined as a selected false detection area.
  • the image frame corresponding to the image to be edited can be added in different ways.
  • an image frame corresponding to at least one image frame coordinate may be determined on the image to be edited. That is, the user can input the coordinates of the image frame representing the position of the image frame to be edited through peripheral devices such as a keyboard and a mouse, and the electronic device directly determines the corresponding image frame when receiving the coordinates of the image frame.
  • a corresponding image frame may be determined according to a position of the at least one preset gesture on the image to be edited. That is to say, the user can also directly draw a corresponding image frame on the image to be edited by dragging with a mouse or dragging with a gesture.
  • the user can add image frames according to each detection frame.
  • the user can also directly select some of the detection frames as image frames.
  • the attribute of the false detection area may be the coordinates of each image frame representing the false detection area.
  • FIG. 3 shows a schematic diagram of a scene display interface 30 according to an embodiment of the present disclosure.
  • the scene display interface 30 can be used to display an edit control 31 , and the edit control 31 is used to enter an edit mode when triggered.
  • the scene display interface 30 can also be used to display the image to be edited 32 and the image information 34 corresponding to the image to be edited 32 .
  • the image to be edited 32 is in an uneditable state, and after the edit control 32 is triggered, the image to be edited 32 enters an editable state.
  • the image to be edited 32 also has a corresponding detection frame 33 , and the detection frame 33 represents an area where an object is detected in the image to be edited 32 after preliminary object detection or historical object detection.
  • the user can select at least one corresponding false detection area based on the marked detection frame 33 in the image 32 to be edited.
  • the electronic device may also directly receive the attributes of false detection areas transmitted by other devices.
  • Step S20 perform object detection on the image to be detected, and obtain at least one object detection result.
  • At least one object detection result is obtained by performing object detection on a predetermined image to be detected.
  • the detection method may be to input the image to be detected into the trained object detection model to perform object detection, and output at least one object detection result.
  • at least one object detection result includes the coordinates of the detection frame and the corresponding probability value, the probability value is used to represent the possibility of the target object in the corresponding detection frame coordinates of the image to be detected, and the detection frame coordinates can be the detection represented by coordinates Region properties.
  • the object detection model will detect the detection area that may be a human face in the image to be detected, and output the coordinates of the detection frame surrounding each detection area, and each detection area includes human faces. The probability value of the face.
  • Fig. 4 shows a schematic diagram of determining an object detection result according to an embodiment of the present disclosure.
  • the image to be detected 40 is input into a pre-trained object detection model 41 , and a corresponding object detection result 42 is output.
  • the content included in the object detection result 42 is determined according to the output format of the object detection model 41 .
  • the detection method may also be inputting the image to be detected into the trained object detection model, and the object detection model extracts image features for object detection, and only determines the obtained multiple The detection frame coordinates of possible objects are used as the object detection results, and the detection results of each object are screened based on the position, and then the probability value detection is further performed to obtain the intrusion detection results.
  • Step S30 Filter the object detection result according to the attribute of the false detection area and the attribute of the detection area to obtain an intrusion detection result.
  • the detection area attribute is used to represent the position and size of the detection area corresponding to the object detection result, and the detection area is the detected area where the object may exist in the image to be detected.
  • the area surrounded by the detection frame is the detection area, and the position and size of the corresponding area can be directly determined according to the coordinates of the detection frame to obtain the attributes of the detection area.
  • the object detection result includes detection frame coordinates and corresponding probability values.
  • the way to determine the intrusion detection result can be to first screen the object detection result according to the corresponding relationship between the false detection area attribute and the detection area attribute, and then according to the probability value in the object detection result, the object detection As a result, the second screening is performed, and the intrusion detection result is determined according to the detection results of each object obtained after the first screening and the second screening.
  • the first screening process may be to delete the detection area attribute corresponding to the object in the object detection result in response to the fact that the detection area corresponding to the detection area attribute is inside a false detection area represented by a false detection area attribute. That is to say, when there is a false detection region attribute corresponding to the positional relationship of the detection region attribute, it is considered that the probability of the current object detection result being a false detection result is high, and the object detection result is deleted.
  • the corresponding positional relationship means that the corresponding detection area is inside or overlaps with the false detection area.
  • the manner of judging whether the detection area is inside the false detection area may be determined by determining the positions of key points in the detection area. For example, the key point position of the object in the object detection result is determined, and in response to the key point position being inside a false detection area represented by a false detection area attribute, the detection area represented by the detection area attribute is determined to be inside the false detection area.
  • the center position of the detection area or the center position of the bottom edge may be set as the key point, and when the key point is inside a false detection area, it is determined that the detection area is inside the false detection area.
  • the second screening process may be to delete the object detection result in response to the probability value in the object detection result being less than a preset threshold. That is to say, when the probability value included in the object detection result is less than the preset threshold, it is considered that the object detection result corresponds to a low probability of an object existing in the detection area, and the object detection result is deleted. For example, when there are three object detection result 1, object detection result 2 and object detection result 3 with probability values of 0.68, 0.59 and 0.81 respectively, and the preset threshold value is 0.6, delete object detection result 2 and keep object detection result Result 1 and object detection result 3.
  • FIG. 5 shows a schematic diagram of a screening object detection result 50 according to an embodiment of the present disclosure.
  • the object detection result 50 includes detection frame coordinates and probability values.
  • the detection area attribute 51 representing the position and size of the detection area is first determined according to the detection frame coordinates in the object detection result. Further, the first screening is performed by comparing whether there is a corresponding relationship between the detection area attribute 51 and each false detection area attribute 52, and the corresponding relationship is whether the represented detection area coincides with the false detection area or is inside the false detection area. If there is a corresponding false detection area attribute 52, the object detection result 50 is directly deleted.
  • the probability value 53 in the detection region attribute is further obtained for the second screening, that is, the magnitude of the probability value 53 and the preset threshold 54 is compared. If the probability value 53 is less than the preset threshold 54, the object detection result 50 is deleted. If the probability value 53 is not less than the preset threshold 54 , an intrusion detection result 55 is further determined according to the object detection result 50 .
  • the execution order of the first screening process and the second screening process can be set randomly, which is not limited here.
  • Fig. 6 shows a schematic diagram of judging the corresponding relationship between false detection area attributes and detection area attributes according to an embodiment of the present disclosure. As shown in FIG. 6 , in the embodiment of the present disclosure, it may be determined whether the attribute of the false detection area corresponds to the attribute of the detection area according to the positional relationship.
  • the embodiment of the present disclosure predetermines two false detection region attributes, and each false detection region attribute corresponds to the first false detection region 62 and the second false detection region 63 as an example for illustration.
  • the detection area attributes corresponding to each object detection result represent the first detection area 60 and the second detection area 61 respectively, by judging the first detection area 60 and the second detection area 61 respectively.
  • the positional relationship between the first false detection area 62 and the second false detection area 63 filters the object detection results.
  • the feature point of the detection area when the feature point of the detection area is within the false detection area, it is judged that the detection area corresponds to the false detection area, and the object detection result corresponding to the detection area is deleted.
  • the feature point can be any point in the detection area, for example, it can be a center point, a vertex of the detection area, and the like.
  • the second detection area 61 in FIG. 6 corresponds to the first false detection area 62 , that is, the object detection result corresponding to the second detection area 61 is deleted.
  • the screening process may also be directly completed in the object detection model of the input image to be detected.
  • the object detection model first extracts image features for object detection, and only determines the obtained detection frame coordinates of multiple possible objects as the object detection result. Further, the object detection model performs screening according to the correspondence between the coordinates of the detection frame in each object detection result and the attributes of each false detection area, and deletes the object detection results with corresponding false detection area attributes. The object detection model further detects the probability values of the objects in the multiple detection frame coordinates obtained after screening, and adds the coordinates and corresponding probability values of the detection frames whose probability values are greater than the preset threshold to the intrusion detection results. Finally, according to the screening results and Each probability value output does not include an object detection result, or an intrusion detection result that includes at least one detection result.
  • the image to be detected is marked according to the intrusion detection result, and the marked image to be detected and corresponding prompt information are output.
  • This method can remind the user of the detection result of this time, and can also be used as an image to be edited in the next detection process to correct the next detection result.
  • the embodiments of the present disclosure can delineate corresponding false detection regions through detected images in specific acquisition scenarios, and filter detection results obtained based on false detection regions during object detection to improve the accuracy of detection results. Furthermore, in scenarios such as product demonstrations, this method can also be used to detect and identify false positive results and correct errors in time.
  • the present disclosure also provides object detection devices, electronic equipment, computer-readable storage media, and programs, all of which can be used to implement any object detection method provided in the present disclosure.
  • object detection devices electronic equipment, computer-readable storage media, and programs, all of which can be used to implement any object detection method provided in the present disclosure.
  • FIG. 7 shows a schematic diagram of an object detection device according to an embodiment of the present disclosure.
  • the object detection device of the embodiment of the present disclosure includes:
  • An attribute determining module 70 configured to determine at least one attribute of the false detection region, the attribute of the false detection region is used to characterize the position and size of the false detection region;
  • the image detection module 71 is configured to perform object detection on the image to be detected to obtain at least one object detection result, wherein the at least one object detection result includes detection area attributes representing the position and size of the detection object;
  • the result screening module 72 is configured to screen the object detection result according to the false detection area attribute and the detection area attribute to obtain an intrusion detection result.
  • the attribute determination module includes:
  • the human-computer interaction sub-module is configured to determine at least one false detection area attribute in response to the false detection area generated by the human-computer interaction operation.
  • the human-computer interaction submodule includes:
  • a pre-detection unit configured to perform object detection on the image to be detected and display a pre-detection result of the object detection, the pre-detection result including at least one preliminary detection frame;
  • the first area determining unit is configured to, in response to receiving at least one preliminary detection frame selected by the user through human-computer interaction, determine the area represented by the preliminary detection frame as a false detection area.
  • the human-computer interaction submodule includes:
  • the edit control display unit is used to display the edit control on the scene display interface
  • An editing image display unit configured to display an image to be edited corresponding to the image to be detected in response to the editing control being triggered;
  • the second area determination unit is configured to, in response to at least one area in the image to be edited being selected, determine that the selected area is a falsely detected area.
  • the second area determination unit includes:
  • an image frame determining subunit configured to determine at least one image frame corresponding to the image to be edited
  • the false detection area determination subunit is configured to determine the area within the at least one image frame as the selected false detection area.
  • the image frame determining subunit includes:
  • a coordinate receiving subunit configured to, in response to receiving at least one image frame coordinate, determine an image frame corresponding to the at least one image frame coordinate on the image to be edited;
  • the gesture determining subunit is configured to, in response to receiving at least one preset gesture corresponding to the image to be edited, project at least one region determined on the image to be edited according to the at least one preset gesture as an image frame.
  • the object detection result includes a detection frame coordinate and a corresponding probability value, and the probability value is used to represent the possibility that the target object exists in the image to be detected within the corresponding detection frame coordinate,
  • the detection frame coordinates are the attributes of the detection area characterized by coordinates.
  • the result screening module includes:
  • the first screening submodule is configured to perform a first screening on the object detection result according to the corresponding relationship between the attribute of the false detection area and the attribute of the detection area;
  • the second screening submodule is used to perform a second screening on the object detection result according to the probability value in the object detection result;
  • the result determination submodule is configured to determine the intrusion detection result according to the object detection results obtained after the first screening and the second screening.
  • the first screening submodule includes;
  • the position screening unit is configured to delete the detection area attribute corresponding to the object in the object detection result in response to the object in the object detection result being located inside the false detection area represented by the false detection area attribute.
  • the location screening unit includes:
  • a key point determining subunit configured to determine the key point position of the object in the object detection result
  • the positional relationship determination subunit is configured to determine that the detection area represented by the attribute of the detection area is inside the false detection area in response to the position of the key point being inside a false detection area represented by a false detection area attribute.
  • the second screening submodule includes:
  • a threshold screening unit configured to delete the object detection result in response to a probability value in the object detection result being less than a preset threshold.
  • the device further includes:
  • An image tagging module configured to tag the image to be detected by the intrusion detection result
  • the information display module is used to output the labeled image to be detected and corresponding prompt information.
  • the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the methods described in the method embodiments above, and its specific implementation can refer to the description of the method embodiments above. For brevity, here No longer.
  • Embodiments of the present disclosure also provide a computer-readable storage medium, on which computer program instructions are stored, and the above-mentioned method is implemented when the computer program instructions are executed by a processor.
  • Computer readable storage media may be volatile or nonvolatile computer readable storage media.
  • An embodiment of the present disclosure also proposes an electronic device, including: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to invoke the instructions stored in the memory to execute the above method.
  • An embodiment of the present disclosure also provides a computer program product, including computer-readable codes, or a non-volatile computer-readable storage medium carrying computer-readable codes, when the computer-readable codes are stored in a processor of an electronic device When running in the electronic device, the processor in the electronic device executes the above method.
  • Electronic devices may be provided as terminals, servers, or other forms of devices.
  • Fig. 8 is a block diagram of an electronic device 800 according to an exemplary embodiment.
  • the electronic device 800 may be a terminal such as a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, or a personal digital assistant.
  • electronic device 800 may include one or more of the following components: processing component 802, memory 804, power supply component 806, multimedia component 808, audio component 810, input/output (I/O) interface 812, sensor component 814 , and the communication component 816.
  • the processing component 802 generally controls the overall operations of the electronic device 800, such as those associated with display, telephone calls, data communications, camera operations, and recording operations.
  • the processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above method. Additionally, processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802 .
  • the memory 804 is configured to store various types of data to support operations at the electronic device 800 . Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phonebook data, messages, pictures, videos, and the like.
  • the memory 804 can be implemented by any type of volatile or non-volatile storage device or their combination, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EPROM erasable Programmable Read Only Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic or Optical Disk Magnetic Disk
  • the power supply component 806 provides power to various components of the electronic device 800 .
  • Power components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for electronic device 800 .
  • the multimedia component 808 includes a screen providing an output interface between the electronic device 800 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user.
  • the touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense a boundary of a touch or swipe action, but also detect duration and pressure associated with the touch or swipe action.
  • the multimedia component 808 includes a front camera and/or a rear camera. When the electronic device 800 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera can receive external multimedia data. Each front camera and rear camera can be a fixed optical lens system or have focal length and optical zoom capability.
  • the audio component 810 is configured to output and/or input audio signals.
  • the audio component 810 includes a microphone (MIC), which is configured to receive external audio signals when the electronic device 800 is in operation modes, such as call mode, recording mode and voice recognition mode. Received audio signals may be further stored in memory 804 or sent via communication component 816 .
  • the audio component 810 also includes a speaker for outputting audio signals.
  • the I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, and the like. These buttons may include, but are not limited to: a home button, volume buttons, start button, and lock button.
  • Sensor assembly 814 includes one or more sensors for providing status assessments of various aspects of electronic device 800 .
  • the sensor component 814 can detect the open/closed state of the electronic device 800, the relative positioning of components, such as the display and the keypad of the electronic device 800, the sensor component 814 can also detect the electronic device 800 or a Changes in position of components, presence or absence of user contact with electronic device 800 , electronic device 800 orientation or acceleration/deceleration and temperature changes in electronic device 800 .
  • Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact.
  • Sensor assembly 814 may also include an optical sensor, such as a complementary metal-oxide-semiconductor (CMOS) or charge-coupled device (CCD) image sensor, for use in imaging applications.
  • CMOS complementary metal-oxide-semiconductor
  • CCD charge-coupled device
  • the sensor component 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.
  • the communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices.
  • the electronic device 800 can access a wireless network based on a communication standard, such as a wireless network (WiFi), a second generation mobile communication technology (2G) or a third generation mobile communication technology (3G), or a combination thereof.
  • the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication.
  • the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID Radio Frequency Identification
  • IrDA Infrared Data Association
  • UWB Ultra Wide Band
  • Bluetooth Bluetooth
  • electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation for performing the methods described above.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable A programmable gate array
  • controller microcontroller, microprocessor or other electronic component implementation for performing the methods described above.
  • a non-volatile computer-readable storage medium such as the memory 804 including computer program instructions, which can be executed by the processor 820 of the electronic device 800 to implement the above method.
  • Fig. 9 is a block diagram of an electronic device 900 according to an exemplary embodiment.
  • the electronic device 900 may be provided as a server.
  • electronic device 900 includes processing component 922 , which further includes one or more processors, and a memory resource represented by memory 932 for storing instructions executable by processing component 922 , such as application programs.
  • the application program stored in memory 932 may include one or more modules each corresponding to a set of instructions.
  • the processing component 922 is configured to execute instructions to perform the above method.
  • Electronic device 900 may also include a power supply component 926 configured to perform power management of electronic device 900, a wired or wireless network interface 950 configured to connect electronic device 900 to a network, and an input-output (I/O) interface 958 .
  • the electronic device 900 can operate based on the operating system stored in the memory 932, such as the Microsoft server operating system (Windows Server TM ), the operating system based on the graphical user interface (Mac OS X TM ) introduced by Apple Inc., and the multi-user and multi-process computer operating system (Unix TM ), a free and open-source Unix-like operating system (Linux TM ), an open-source Unix-like operating system (FreeBSD TM ), or the like.
  • Microsoft server operating system Windows Server TM
  • Mac OS X TM graphical user interface
  • Uniix TM multi-user and multi-process computer operating system
  • FreeBSD TM open-source Unix-like operating system
  • a non-transitory computer-readable storage medium such as the memory 932 including computer program instructions, which can be executed by the processing component 922 of the electronic device 900 to implement the above method.
  • the present disclosure can be a system, method and/or computer program product.
  • a computer program product may include a computer readable storage medium having computer readable program instructions thereon for causing a processor to implement various aspects of the present disclosure.
  • a computer readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device.
  • a computer readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing.
  • Computer-readable storage media include: portable computer diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory), static random access memory (SRAM), compact disc read only memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanically encoded device, such as a printer with instructions stored thereon A hole card or a raised structure in a groove, and any suitable combination of the above.
  • RAM random access memory
  • ROM read-only memory
  • EPROM erasable programmable read-only memory
  • flash memory static random access memory
  • SRAM static random access memory
  • CD-ROM compact disc read only memory
  • DVD digital versatile disc
  • memory stick floppy disk
  • mechanically encoded device such as a printer with instructions stored thereon
  • a hole card or a raised structure in a groove and any suitable combination of the above.
  • computer-readable storage media are not to be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., pulses of light through fiber optic cables), or transmitted electrical signals.
  • the computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or downloaded to an external computer or external storage device over a network, such as the Internet, local area network, wide area network, and/or wireless network.
  • the network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers.
  • a network adapter card or a network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing/processing device .
  • Computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or Source or object code written in any combination, including object-oriented programming languages—such as Smalltalk, C++, etc., and conventional procedural programming languages—such as the “C” language or similar programming languages.
  • Computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server implement.
  • the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (such as via the Internet using an Internet service provider). connect).
  • LAN local area network
  • WAN wide area network
  • an electronic circuit such as a programmable logic circuit, field programmable gate array (FPGA), or programmable logic array (PLA)
  • FPGA field programmable gate array
  • PDA programmable logic array
  • These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine such that when executed by the processor of the computer or other programmable data processing apparatus , producing an apparatus for realizing the functions/actions specified in one or more blocks in the flowchart and/or block diagram.
  • These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause computers, programmable data processing devices and/or other devices to work in a specific way, so that the computer-readable medium storing instructions includes An article of manufacture comprising instructions for implementing various aspects of the functions/acts specified in one or more blocks in flowcharts and/or block diagrams.
  • each block in a flowchart or block diagram may represent a module, a portion of a program segment, or an instruction that includes one or more Executable instructions.
  • the functions noted in the block may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations can be implemented by a dedicated hardware-based system that performs the specified function or action , or may be implemented by a combination of dedicated hardware and computer instructions.
  • the computer program product can be specifically realized by means of hardware, software or a combination thereof.
  • the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK) etc. wait.
  • a software development kit Software Development Kit, SDK

Landscapes

  • Image Analysis (AREA)

Abstract

本公开涉及一种对象检测方法及装置、电子设备和存储介质，通过确定至少一个用于表征误检测区域位置和尺寸的误检区域属性，对待检测图像进行对象检测得到至少一个对象检测结果，其中，至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性。根据误检区域属性和检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。本公开实施例能够通过划定特定场景下的误检测区域，在进行对象检测时基于误检测区域筛选得到的检测结果以提升检测结果的准确性。进一步地，在进行产品演示等场景下，还能够通过该方法检测识别误报结果并及时更正错误。

Description

对象检测方法及装置、电子设备和存储介质
本申请要求2021年09月27日提交、申请号为202111135692.7，发明名称为“对象检测方法及装置、电子设备和存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本公开涉及计算机技术领域，尤其涉及一种对象检测方法及装置、电子设备和存储介质。
背景技术
在进行人脸识别、车牌检测等对象检测场景下，经常会产生误报的情况。即在没有相应对象的位置识别到了对象。
发明内容
本公开提出了一种对象检测方法及装置、电子设备和存储介质，能够在不更正算法的情况下及时有效的修正检测结果，提高检测结果的准确性。
根据本公开的第一方面，提供了一种对象检测方法，所述方法包括：
确定至少一个误检区域属性，所述误检区域属性用于表征误检测区域的位置和尺寸；
对待检测图像进行对象检测，得到至少一个对象检测结果，其中，所述至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性；
根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
在一种可能的实现方式中，所述确定至少一个误检区域属性包括：
响应于人机交互操作生成的误检测区域，确定至少一个误检测区域属性。
在一种可能的实现方式中，所述响应于人机交互操作生成的误检测区域，通过以下方式实现：
对所述待检测图像进行对象检测并显示所述对象检测的预检测结果，所述预检测结果中包括至少一个初步检测框；
响应于接收到用户通过人机交互操作选中的至少一个初步检测框，确定所述初步检测框表征的区域为误检测区域。
在一种可能的实现方式中，所述响应于人机交互操作生成的误检测区域，通过以下方式实现：
在场景显示界面显示编辑控件；
响应于所述编辑控件被触发，显示与所述待检测图像对应的待编辑图像；
响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域。
在一种可能的实现方式中，所述响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域包括：
确定所述待编辑图像对应的至少一个图像框；
确定至少一个所述图像框内的区域为被选中的误检测区域。
在一种可能的实现方式中，所述确定所述待编辑图像对应的至少一个图像框包括：
响应于接收到至少一个图像框坐标，在所述待编辑图像上确定至少一个所述图像框坐标对应的图像框；或者
响应于接收到对应于所述待编辑图像的至少一个预设手势，根据至少一个所述预设手势投影在所述待编辑图像上确定的至少一个区域作为图像框。
在一种可能的实现方式中，所述对象检测结果中包括检测框坐标和对应的概率值，所述概率值用于表征所述待检测图像在对应检测框坐标内存在目标对象的可能性，所述检测框坐标为通过坐标方式表征的检测区域属性。
在一种可能的实现方式中，所述根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果包括：
根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选；
根据所述对象检测结果中的概率值，对所述对象检测结果进行第二次筛选；
根据第一次筛选和第二次筛选后得到的对象检测结果确定入侵检测结果。
在一种可能的实现方式中，所述根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选包括；
响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，删除所述对象检测结果中的所述对象对应的检测区域属性。
在一种可能的实现方式中，所述响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，通过以下方式实现：
确定所述对象检测结果中的对象的关键点位置；
响应于所述关键点位置在一个误检区域属性表征的误检区域内部，确定所述检测区域属性表征得到检测区域在所述误检区域内部。
在一种可能的实现方式中，所述根据所述对象检测结果中的概率值进行第二次筛选包括：
响应于对象检测结果中的概率值小于预设阈值，删除所述对象检测结果。
在一种可能的实现方式中，所述方法还包括：
通过所述入侵检测结果标注所述待检测图像；
输出标注后的待检测图像和对应的提示信息。
根据本公开的第二方面，提供了一种对象检测装置，所述装置包括：
属性确定模块，用于确定至少一个误检区域属性，所述误检区域属性用于表征误检测区域的位置和尺寸；
图像检测模块，用于对待检测图像进行对象检测，得到至少一个对象检测结果，其中，所述至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性；
结果筛选模块，用于根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
在一种可能的实现方式中，所述属性确定模块包括：
人机交互子模块，用于响应于人机交互操作生成的误检测区域，确定至少一个误检测区域属性。
在一种可能的实现方式中，所述人机交互子模块包括：
预检测单元，用于对所述待检测图像进行对象检测并显示所述对象检测的预检测结果，所述预检测结果中包括至少一个初步检测框；
第一区域确定单元，用于响应于接收到用户通过人机交互操作选中的至少一个初步检测框，确定所述初步检测框表征的区域为误检测区域。
在一种可能的实现方式中，所述人机交互子模块包括：
编辑控件显示单元，用于在场景显示界面显示编辑控件；
编辑图像显示单元，用于响应于所述编辑控件被触发，显示与所述待检测图像对应的待编辑图像；
第二区域确定单元，用于响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域。
在一种可能的实现方式中，所述第二区域确定单元包括：
图像框确定子单元，用于确定所述待编辑图像对应的至少一个图像框；
误检区域确定子单元，用于确定所述至少一个图像框内的区域为被选中的误检测区域。
在一种可能的实现方式中，所述图像框确定子单元包括：
坐标接收子单元，用于响应于接收到至少一个图像框坐标，在所述待编辑图像上确定所述至少一个图像框坐标对应的图像框；或者
手势确定子单元，用于响应于接收到对应于所述待编辑图像的至少一个预设手势，根据所述至少一个预设手势投影在所述待编辑图像上确定的至少一个区域作为图像框。
在一种可能的实现方式中，所述对象检测结果中包括检测框坐标和对应的概率值，所述概率值用于表征所述待检测图像在对应检测框坐标内存在目标对象的可能性，所述检测框坐标为通过坐标方式表征的检测区域属性。
在一种可能的实现方式中，所述结果筛选模块包括：
第一筛选子模块，用于根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选；
第二筛选子模块，用于根据所述对象检测结果中的概率值，对所述对象检测结果进行第二次筛选；
结果确定子模块，用于根据第一次筛选和第二次筛选后得到的对象检测结果确定入侵检测结果。
在一种可能的实现方式中，所述第一筛选子模块包括；
位置筛选单元，用于响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，删除所述对象检测结果中的所述对象对应的检测区域属性。
在一种可能的实现方式中，所述位置筛选单元包括：
关键点确定子单元，用于确定所述对象检测结果中的对象的关键点位置；
位置关系确定子单元，用于响应于所述关键点位置在一个误检区域属性表征的误检区域内部，确定所述检测区域属性表征得到检测区域在所述误检区域内部。
在一种可能的实现方式中，所述第二筛选子模块包括：
阈值筛选单元，用于响应于对象检测结果中的概率值小于预设阈值，删除所述对象检测结果。
在一种可能的实现方式中，所述装置还包括：
图像标注模块，用于通过所述入侵检测结果标注所述待检测图像；
信息显示模块，用于输出标注后的待检测图像和对应的提示信息。
根据本公开的第三方面，提供了一种电子设备，包括：处理器；用于存储处理器可执行指令的存 储器；其中，所述处理器被配置为调用所述存储器存储的指令，以执行上述方法。
根据本公开的第四方面，提供了一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述方法。
根据本公开的第五方面，提供了一种计算机程序产品，包括计算机可读代码，或者承载有计算机可读代码的非易失性计算机可读存储介质，当所述计算机可读代码在电子设备的处理器中运行时，所述电子设备中的处理器执行用于实现上述方法。
在本公开实施例中，该对象检测方法通过划定特定场景下的误检测区域，在进行对象检测时基于误检测区域筛选得到的检测结果以提升检测结果的准确性。进一步地，在进行产品演示等场景下，还能够通过该方法检测识别误报结果并及时更正错误。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，而非限制本公开。根据下面参考附图对示例性实施例的详细说明，本公开的其它特征及方面将变得清楚。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，这些附图示出了符合本公开的实施例，并与说明书一起用于说明本公开的技术方案。
图1示出根据本公开实施例的一种对象检测方法的流程图；
图2示出根据本公开实施例的确定误检区域属性过程的流程图；
图3示出根据本公开实施例的一种场景显示界面的示意图；
图4示出根据本公开实施例的一种确定对象检测结果的示意图；
图5示出根据本公开实施例的一种筛选对象检测结果的示意图；
图6示出根据本公开实施例的判断误检区域属性和检测区域属性对应关系的示意图；
图7示出根据本公开实施例的对象检测装置的示意图；
图8是根据示例性实施例示出的一种电子设备的框图；
图9是根据示例性实施例示出的另一种电子设备的框图。
具体实施方式
以下将参考附图详细说明本公开的各种示例性实施例、特征和方面。附图中相同的附图标记表示功能相同或相似的元件。尽管在附图中示出了实施例的各种方面，但是除非特别指出，不必按比例绘制附图。
在这里专用的词“示例性”意为“用作例子、实施例或说明性”。这里作为“示例性”所说明的任何实施例不必解释为优于或好于其它实施例。
本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中术语“至少一种”表示多种中的任意一种或多种中的至少两种的任意组合，例如，包括A、B、C中的至少一种，可以表示包括从A、B和C构成的集合中选择的任意一个或多个元素。
另外，为了更好地说明本公开，在下文的具体实施方式中给出了众多的具体细节。本领域技术人 员应当理解，没有某些具体细节，本公开同样可以实施。在一些实例中，对于本领域技术人员熟知的方法、手段、元件和电路未作详细描述，以便于凸显本公开的主旨。
图1示出根据本公开实施例的一种对象检测方法的流程图。在一种可能的实现方式中，该对象检测方法可以由终端设备或服务器等电子设备执行，终端设备可以为用户设备(User Equipment，UE)、移动设备、用户终端、终端、蜂窝电话、无绳电话、个人数字助理(Personal Digital Assistant，PDA)、手持设备、计算设备、车载设备、可穿戴设备等，所述方法步骤的执行主体可以为硬件执行，或者通过处理器运行计算机可执行代码的方式执行。例如，可以通过处理器调用存储器中存储的计算机可读指令的方式来实现。或者，可通过服务器执行所述方法。其中，服务器可以为单个的服务器，或者由多个服务器组成的服务器集群。
在一个示例性的应用场景中，可根据历史对象检测过程中得到的检测结果，利用终端设备或其它处理设备执行本公开实施例的对象检测方法，来修正对象检测过程的误差，从而提高对象检测结果的准确程度。
如图1所示，本公开实施例的对象检测方法包括以下步骤：
步骤S10、确定至少一个误检区域属性。
在一种可能的实现方式中，误检区域属性用于表征误检测区域的位置和尺寸，例如可以包括用于划定误检测区域的图像框坐标，该图像框内的区域为误检测区域。可选地，该误检测区域为在检测过程中，有较大可能在不存在对象的情况下误检测到对象的区域。误检测区域可以基于历史的对象检测结果或者待检测图像的初步检测结果确定。
在一个可选的应用场景中，本公开实施例的对象检测方法用于对固定场景的对象检测。也就是说，在同一场景下的特定区域经常出现错误检测的情况下，将该区域确定为误检测区域，以在后续的对象检测过程中通过误检测区域的误检区域属性筛选检测结果，在不需要修改底层算法的情况下提高检测结果的准确程度。
在另一个可选的应用场景中，本公开实施例的对象检测方法用于对特定待检测图像的检测结果进行修正。也就是说，在对待检测图像进行正式的对象检测之前，先进行一次初步的对象检测。在该初步对象检测的结果中出现错误检测的情况下，将该错误检测区域确定为误检测区域，以在正式的对象检测过程中通过误检测区域的误检区域属性筛选检测结果，提高正式检测结果的准确程度。
可选地，在上述任意应用场景下，误检测区域属性的确定方式都可以为响应于人机交互操作生成的误检测区域，确定至少一个误检测区域属性。也就是说，用户可以直接通过鼠标、键盘、触摸屏等电子设备的人机交互模块与电子设备进行人机交互，以根据历史的对象检测几个或者待检测图像的初步检测结果选中至少一个误检测区域得到对应的误检区域属性。
在一种可能的实现方式中，本公开实施例响应于人机交互操作生成误检测区域的过程还可以包括：对待检测图像进行对象检测并显示对象检测的预检测结果，预检测结果中包括至少一个初步检测框。响应于接收到用户通过人机交互操作选中的至少一个初步检测框，确定初步检测框表征的区域为误检测区域。也就是说，可以对待检测图像进行正式的对象检测之前，先进行一次初步的对象检测得到预检测结果，预检测结果中包括至少一个初步检测框。电子设备通过显示装置显示带有初步检测框的待检测图像，用户可以直接识别各初步检测框中是否存在对象，不存在的情况下直接选中该初步检测框， 以将该初步检测框内的区域作为误检测区域。
图2示出根据本公开实施例的确定误检区域属性过程的流程图。如图2所示，本公开实施例通过人机交互方式确定误检区域属性的过程包括以下步骤：
步骤S11、在场景显示界面显示编辑控件。
在一种可能的实现方式中，该场景显示界面用于显示编辑控件。该编辑控件用于在被触发的情况下进入编辑模式，以通过编辑选中误检测区域。可选地，该场景显示界面还用于显示待检测图像对应的待编辑图像、以及该待编辑图像对应的图像信息。其中，该待编辑图像对应于本次对象检测过程的待检测图像，用户可以在触发编辑控件后基于该待编辑图像选中误检测区域。
进一步地，场景显示界面中还可以显示用于替换、增加、删除待编辑图像的图像修改控件等内容。
步骤S12、响应于所述编辑控件被触发，显示与所述待检测图像对应的待编辑图像。
在一种可能的实现方式中，用户可以通过预设手势点击电子设备的触摸屏，以及通过鼠标单击或双击等方式触发该编辑控件。在编辑控件被触发时，进入编辑模式，即在场景显示界面显示待编辑图像。其中，在场景显示界面中已经显示待编辑图像的情况下，不再重新显示该图像，直接将该待编辑图像由不可编辑状态转换为可编辑状态。在场景显示界面中未显示待编辑图像的情况下，载入并显示与待检测图像对应的待编辑图像，并将该待编辑图像的状态设定为可编辑状态。
进一步地，待编辑图像可以为与需要进行对象检测的待检测图像采集场景相同的图像，可以根据不同的应用场景选取。例如，在对经过初步对象检测的图像再次进行正式对象检测的应用场景下，可以直接确定待检测图像为待编辑图像。在根据历史图像检测结果修正待检测图像的对象检测结果的应用场景下，可以获取同一采集场景采集到，且已经完成对象检测的图像作为待编辑图像。可选地，采集场景可以为楼梯间、街道、走廊、小区内等场景。该待编辑图像上还显示经过历史对象检测，或者初步对象检测后检测到的对象所在位置，该位置可以通过对应的检测框标注。
步骤S13、响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域。
在一种可能的实现方式中，通过选中待编辑图像中部分区域的方式确定误检测区域。可选地，该待编辑图像中的部分区域通过在待编辑图像上添加图像框的方式选中。也就是说，确定待编辑图像对应的至少一个图像框，确定至少一个图像框内的区域为被选中的误检测区域。
进一步地，待编辑图像对应的图像框可以通过不同方式添加。例如，可以响应于接收到至少一个图像框坐标，在所述待编辑图像上确定至少一个图像框坐标对应的图像框。即用户可以通过键盘、鼠标等外设输入表征图像框在待编辑图像位置的图像框坐标，电子设备在接收到图像框坐标时直接确定对应的图像框。或者，还可以响应于接收到对应于所述待编辑图像的至少一个预设手势，根据至少一个预设手势在待编辑图像上的位置确定对应的图像框。也就是说，用户还可以通过鼠标拖动，或手势拖动的方式在待编辑图像上直接绘制对应的图像框。
可选地，在该待编辑图像中还包括经过初步对象检测或历史对象检测得到的检测框时，用户可以根据各检测框添加图像框。或者，用户还可以直接选择其中部分检测框作为图像框。
在一个可选的实现方式中，在确定待编辑图像中的至少一个误检测区域后，根据各误检测区域在待编辑图像中所处的位置，以及待检测区域的尺寸确定对应的误检区域属性。可选地，该误检区域属性可以为表征误检测区域的各图像框坐标。
图3示出根据本公开实施例的一种场景显示界面30的示意图。如图3所示，该场景显示界面30可以用于显示编辑控件31，该编辑控件31用于在被触发的情况下进入编辑模式。进一步地，该场景显示界面30还可以用于显示待编辑图像32和待编辑图像32对应的图像信息34。在编辑控件31被触发之前，待编辑图像32处于不可编辑状态，在编辑控件32被触发之后，待编辑图像32进入编辑状态。可选地，该待编辑图像32上还具有对应的检测框33，该检测框33表征待编辑图像32在经过初步对象检测或历史对象检测后，检测到对象的区域。在进入编辑模式后，用户可以基于待编辑图像32中标注的检测框33选定对应的至少一个误检测区域。
进一步地，电子设备还可以直接接收其他设备传输的误检区域属性。
步骤S20、对待检测图像进行对象检测，得到至少一个对象检测结果。
在一种可能的实现方式中，对预先确定的待检测图像进行对象检测得到至少一个对象检测结果。该检测方式可以为将待检测图像输入训练得到的对象检测模型进行对象检测，输出至少一个对象检测结果。其中，至少一个对象检测结果中包括检测框坐标和对应的概率值，概率值用于表征待检测图像在对应检测框坐标内存在目标对象的可能性，检测框坐标可以为通过坐标方式表征的检测区域属性。例如，在本公开实施例中的对象检测为人脸识别时，对象检测模型会检测待检测图像中可能为人脸的检测区域，并输出包围各检测区域的检测框坐标，以及各检测区域中包括人脸的概率值。
图4示出根据本公开实施例的一种确定对象检测结果的示意图。如图4所示，本公开实施例在确定待检测图像40后，将该待检测图像40输入预先训练得到的对象检测模型41中，输出对应的对象检测结果42。其中，对象检测结果42中包括的内容根据该对象检测模型41的输出格式确定。
在本公开实施例的另一种可能的实现方式中，该检测方式还可以为将待检测图像输入训练得到的对象检测模型中，由对象检测模型提取图像特征进行对象检测，仅确定得到的多个可能存在对象的检测框坐标作为对象检测结果，以基于位置对各对象检测结果筛选后进一步再进行概率值检测得到入侵检测结果。
步骤S30、根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
在一种可能的实现方式中，该检测区域属性用于表征对象检测结果对应检测区域的位置和尺寸，该检测区域为检测得到待检测图像中可能存在对象的区域，即检测区域属性表征待检测图像中可能存在对象的区域位置和尺寸。可选地，当各对象检测结果中包括检测框坐标时，该检测框包围的区域即为检测区域，可以直接根据检测框坐标确定对应区域的位置和尺寸，以得到检测区域属性。
在一种可能的实现方式中，对象检测结果中包括检测框坐标和对应的概率值。在这种情况下，入侵检测结果的确定方式可以为根据误检区域属性与检测区域属性的对应关系，对对象检测结果进行第一次筛选，再根据对象检测结果中的概率值，对对象检测结果进行第二次筛选，根据第一次筛选和第二次筛选后得到的各对象检测结果确定入侵检测结果。
可选地，第一次筛选过程可以为响应于检测区域属性对应检测区域在一个误检区域属性表征的误检测区域内部，删除对象检测结果中的所述对象对应的检测区域属性。也就是说，当存在与检测区域属性位置关系相对应的误检区域属性时，认为当前对象检测结果为误检测结果的概率较大，删除该对象检测结果。其中，位置关系相对应即对应的检测区域在误检测区域内部或与误检测区域重合。
可选地，检测区域是否在误检测区域内部的判断方式可以通过确定检测区域中关键点的位置判断。例如，确定对象检测结果中的对象的关键点位置，响应于关键点位置在一个误检区域属性表征的误检区域内部，确定检测区域属性表征得到检测区域在误检区域内部。在一种可能的实现方式中，可以设定检测区域中心位置或底边中心位置为关键点，当关键点处于一个误检测区域内部时，确定检测区域在该误检测区域内部。
进一步地，第二次筛选过程可以为响应于对象检测结果中的概率值小于预设阈值，删除对象检测结果。也就是说，在对象检测结果中包括的概率值小于预设的阈值时，认为该对象检测结果对应检测区域中存在对象的概率较小，删除该对象检测结果。例如，当存在三个概率值分别为0.68、0.59和0.81的对象检测结果1、对象检测结果2和对象检测结果3，且预先设定的阈值为0.6时，删除对象检测结果2，保留对象检测结果1和对象检测结果3。
图5示出根据本公开实施例的一种筛选对象检测结果50的示意图。如图5所示，在一种可能的实现方式中，对象检测结果50中包括检测框坐标和概率值。在确定对象检测结果50后，先根据该对象检测结果中的检测框坐标确定表征检测区域位置和尺寸的检测区域属性51。进一步地，通过对比该检测区域属性51和各误检区域属性52是否存在对应关系进行第一次筛选，该对应关系即表征的检测区域是否与误检测区域重合，或在误检测区域内部。在存在对应的误检区域属性52的情况下，直接删除该对象检测结果50。在不存在对应的误检区域属性52的情况下，进一步获取该检测区域属性中的概率值53进行第二次筛选，即对比概率值53和预设阈值54的大小。在概率值53小于预设阈值54的情况下，删除该对象检测结果50。在概率值53不小于预设阈值54的情况下，根据对象检测结果50进一步确定入侵检测结果55。
可选地，该第一次筛选过程和第二次筛选过程的执行顺序可以随机设定，在此不做限制。
图6示出根据本公开实施例的判断误检区域属性和检测区域属性对应关系的示意图。如图6所示，本公开实施例可以根据位置关系判断误检区域属性是否和检测区域属性对应。
以本公开实施例预先确定两个误检区域属性，各误检区域属性分别对应第一误检测区域62和第二误检测区域63为例进行说明。当对象检测后得到两个对象检测结果，各对象检测结果对应的检测区域属性分别表征第一检测区域60和第二检测区域61时，通过分别判断第一检测区域60和第二检测区域61与第一误检测区域62和第二误检测区域63的位置关系筛选对象检测结果。
进一步地，当检测区域的特征点在误检测区域内时，即判断该检测区域与误检测区域对应，删除该检测区域对应的对象检测结果。该特征点可以为检测区域中任意点，例如可以是中心点、检测区域的顶点等。例如，当检测区域的特征点为该检测区域底边的中点时，图6中的第二检测区域61与第一误检测区域62对应，即删除第二检测区域61对应的对象检测结果。
在本公开实施例的另一种可能的实现方式中，该筛选过程还可以在待检测图像输入的对象检测模型中直接完成。可选地，在将待检测图像输入训练得到的对象检测模型后，先由对象检测模型提取图像特征进行对象检测，仅确定得到的多个可能存在对象的检测框坐标作为对象检测结果。进一步地，对象检测模型根据各对象检测结果中的检测框坐标与各误检区域属性的对应关系进行筛选，删除存在对应误检区域属性的对象检测结果。对象检测模型再进一步检测筛选后得到的多个检测框坐标内包括对象的概率值，并将其中概率值大于预设阈值的检测框坐标和对应的概率值加入入侵检测结果，最终 根据筛选结果和各概率值输出不包括对象检测结果，或者包括至少一个检测结果的入侵检测结果。
进一步地，在得到入侵检测结果后，通过入侵检测结果标注待检测图像，并输出标注后的待检测图像和对应的提示信息。该方式可以提示用户本次检测结果，同时还可以作为下一次检测过程的待编辑图像，以修正下一次检测结果。
本公开实施例能够通过特定采集场景下已经检测过的图像划定对应的误检测区域，在进行对象检测时基于误检测区域筛选得到的检测结果以提升检测结果的准确性。进一步地，在进行产品演示等场景下，还能够通过该方法检测识别误报结果并及时更正错误。
可以理解，本公开提及的上述各个方法实施例，在不违背原理逻辑的情况下，均可以彼此相互结合形成结合后的实施例，限于篇幅，本公开不再赘述。本领域技术人员可以理解，在具体实施方式的上述方法中，各步骤的具体执行顺序应当以其功能和可能的内在逻辑确定。
此外，本公开还提供了对象检测装置、电子设备、计算机可读存储介质、程序，上述均可用来实现本公开提供的任一种对象检测方法，相应技术方案和描述和参见方法部分的相应记载，不再赘述。
图7示出根据本公开实施例的对象检测装置的示意图，如图7所示，本公开实施例的对象检测装置包括：
属性确定模块70，用于确定至少一个误检区域属性，所述误检区域属性用于表征误检测区域的位置和尺寸；
图像检测模块71，用于对待检测图像进行对象检测，得到至少一个对象检测结果，其中，所述至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性；
结果筛选模块72，用于根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
在一种可能的实现方式中，所述属性确定模块包括：
人机交互子模块，用于响应于人机交互操作生成的误检测区域，确定至少一个误检测区域属性。
在一种可能的实现方式中，所述人机交互子模块包括：
预检测单元，用于对所述待检测图像进行对象检测并显示所述对象检测的预检测结果，所述预检测结果中包括至少一个初步检测框；
第一区域确定单元，用于响应于接收到用户通过人机交互操作选中的至少一个初步检测框，确定所述初步检测框表征的区域为误检测区域。
在一种可能的实现方式中，所述人机交互子模块包括：
编辑控件显示单元，用于在场景显示界面显示编辑控件；
编辑图像显示单元，用于响应于所述编辑控件被触发，显示与所述待检测图像对应的待编辑图像；
第二区域确定单元，用于响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域。
在一种可能的实现方式中，所述第二区域确定单元包括：
图像框确定子单元，用于确定所述待编辑图像对应的至少一个图像框；
误检区域确定子单元，用于确定所述至少一个图像框内的区域为被选中的误检测区域。
在一种可能的实现方式中，所述图像框确定子单元包括：
坐标接收子单元，用于响应于接收到至少一个图像框坐标，在所述待编辑图像上确定所述至少一个图像框坐标对应的图像框；或者
手势确定子单元，用于响应于接收到对应于所述待编辑图像的至少一个预设手势，根据所述至少一个预设手势投影在所述待编辑图像上确定的至少一个区域作为图像框。
在一种可能的实现方式中，所述对象检测结果中包括检测框坐标和对应的概率值，所述概率值用于表征所述待检测图像在对应检测框坐标内存在目标对象的可能性，所述检测框坐标为通过坐标方式表征的检测区域属性。
在一种可能的实现方式中，所述结果筛选模块包括：
第一筛选子模块，用于根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选；
第二筛选子模块，用于根据所述对象检测结果中的概率值，对所述对象检测结果进行第二次筛选；
结果确定子模块，用于根据第一次筛选和第二次筛选后得到的对象检测结果确定入侵检测结果。
在一种可能的实现方式中，所述第一筛选子模块包括；
位置筛选单元，用于响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，删除所述对象检测结果中的所述对象对应的检测区域属性。
在一种可能的实现方式中，所述位置筛选单元包括：
关键点确定子单元，用于确定所述对象检测结果中的对象的关键点位置；
位置关系确定子单元，用于响应于所述关键点位置在一个误检区域属性表征的误检区域内部，确定所述检测区域属性表征得到检测区域在所述误检区域内部。
在一种可能的实现方式中，所述第二筛选子模块包括：
阈值筛选单元，用于响应于对象检测结果中的概率值小于预设阈值，删除所述对象检测结果。
在一种可能的实现方式中，所述装置还包括：
图像标注模块，用于通过所述入侵检测结果标注所述待检测图像；
信息显示模块，用于输出标注后的待检测图像和对应的提示信息。
在一些实施例中，本公开实施例提供的装置具有的功能或包含的模块可以用于执行上文方法实施例描述的方法，其具体实现可以参照上文方法实施例的描述，为了简洁，这里不再赘述。
本公开实施例还提出一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述方法。计算机可读存储介质可以是易失性或非易失性计算机可读存储介质。
本公开实施例还提出一种电子设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为调用所述存储器存储的指令，以执行上述方法。
本公开实施例还提供了一种计算机程序产品，包括计算机可读代码，或者承载有计算机可读代码的非易失性计算机可读存储介质，当所述计算机可读代码在电子设备的处理器中运行时，所述电子设备中的处理器执行上述方法。
电子设备可以被提供为终端、服务器或其它形态的设备。
图8是根据示例性实施例示出的一种电子设备800的框图。例如，电子设备800可以是移动电话， 计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等终端。
参照图8，电子设备800可以包括以下一个或多个组件：处理组件802，存储器804，电源组件806，多媒体组件808，音频组件810，输入/输出(I/O)的接口812，传感器组件814，以及通信组件816。
处理组件802通常控制电子设备800的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件802可以包括一个或多个处理器820来执行指令，以完成上述的方法的全部或部分步骤。此外，处理组件802可以包括一个或多个模块，便于处理组件802和其他组件之间的交互。例如，处理组件802可以包括多媒体模块，以方便多媒体组件808和处理组件802之间的交互。
存储器804被配置为存储各种类型的数据以支持在电子设备800的操作。这些数据的示例包括用于在电子设备800上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器804可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件806为电子设备800的各种组件提供电力。电源组件806可以包括电源管理系统，一个或多个电源，及其他与为电子设备800生成、管理和分配电力相关联的组件。
多媒体组件808包括在所述电子设备800和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件808包括一个前置摄像头和/或后置摄像头。当电子设备800处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件810被配置为输出和/或输入音频信号。例如，音频组件810包括一个麦克风(MIC)，当电子设备800处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器804或经由通信组件816发送。在一些实施例中，音频组件810还包括一个扬声器，用于输出音频信号。
I/O接口812为处理组件802和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件814包括一个或多个传感器，用于为电子设备800提供各个方面的状态评估。例如，传感器组件814可以检测到电子设备800的打开/关闭状态，组件的相对定位，例如所述组件为电子设备800的显示器和小键盘，传感器组件814还可以检测电子设备800或电子设备800一个组件的位置改变，用户与电子设备800接触的存在或不存在，电子设备800方位或加速/减速和电子设备800的温度变化。传感器组件814可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件814还可以包括光传感器，如互补金属氧化物半导体(CMOS)或电荷耦合装置(CCD)图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件814还可以包括加速度传感器， 陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件816被配置为便于电子设备800和其他设备之间有线或无线方式的通信。电子设备800可以接入基于通信标准的无线网络，如无线网络(WiFi)，第二代移动通信技术(2G)或第三代移动通信技术(3G)，或它们的组合。在一个示例性实施例中，通信组件816经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件816还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，电子设备800可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器804，上述计算机程序指令可由电子设备800的处理器820执行以完成上述方法。
图9是根据示例性实施例示出的一种电子设备900的框图。例如，电子设备900可以被提供为一服务器。参照图9，电子设备900包括处理组件922，其进一步包括一个或多个处理器，以及由存储器932所代表的存储器资源，用于存储可由处理组件922的执行的指令，例如应用程序。存储器932中存储的应用程序可以包括一个或一个以上的每一个对应于一组指令的模块。此外，处理组件922被配置为执行指令，以执行上述方法。
电子设备900还可以包括一个电源组件926被配置为执行电子设备900的电源管理，一个有线或无线网络接口950被配置为将电子设备900连接到网络，和一个输入输出(I/O)接口958。电子设备900可以操作基于存储在存储器932的操作系统，例如微软服务器操作系统(Windows Server TM)，苹果公司推出的基于图形用户界面操作系统(Mac OS X TM)，多用户多进程的计算机操作系统(Unix TM),自由和开放原代码的类Unix操作系统(Linux TM)，开放原代码的类Unix操作系统(FreeBSD TM)或类似。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器932，上述计算机程序指令可由电子设备900的处理组件922执行以完成上述方法。
本公开可以是系统、方法和/或计算机程序产品。计算机程序产品可以包括计算机可读存储介质，其上载有用于使处理器实现本公开的各个方面的计算机可读程序指令。
计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是(但不限于)电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身，诸如无线电波或者其他自由传播的电磁波、通过波导或其他传输媒介传播的电磁波(例如，通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
这里所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备，或者 通过网络、例如因特网、局域网、广域网和/或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令，并转发该计算机可读程序指令，以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本公开操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码，所述编程语言包括面向对象的编程语言—诸如Smalltalk、C++等，以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中，通过利用计算机可读程序指令的状态信息来个性化定制电子电路，例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA)，该电子电路可以执行计算机可读程序指令，从而实现本公开的各个方面。
这里参照根据本公开实施例的方法、装置(系统)和计算机程序产品的流程图和/或框图描述了本公开的各个方面。应当理解，流程图和/或框图的每个方框以及流程图和/或框图中各方框的组合，都可以由计算机可读程序指令实现。
这些计算机可读程序指令可以提供给通用计算机、专用计算机或其它可编程数据处理装置的处理器，从而生产出一种机器，使得这些指令在通过计算机或其它可编程数据处理装置的处理器执行时，产生了实现流程图和/或框图中的一个或多个方框中规定的功能/动作的装置。也可以把这些计算机可读程序指令存储在计算机可读存储介质中，这些指令使得计算机、可编程数据处理装置和/或其他设备以特定方式工作，从而，存储有指令的计算机可读介质则包括一个制造品，其包括实现流程图和/或框图中的一个或多个方框中规定的功能/动作的各个方面的指令。
也可以把计算机可读程序指令加载到计算机、其它可编程数据处理装置、或其它设备上，使得在计算机、其它可编程数据处理装置或其它设备上执行一系列操作步骤，以产生计算机实现的过程，从而使得在计算机、其它可编程数据处理装置、或其它设备上执行的指令实现流程图和/或框图中的一个或多个方框中规定的功能/动作。
附图中的流程图和框图显示了根据本公开的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段或指令的一部分，所述模块、程序段或指令的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个连续的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或动作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
该计算机程序产品可以具体通过硬件、软件或其结合的方式实现。在一个可选实施例中，所述计 算机程序产品具体体现为计算机存储介质，在另一个可选实施例中，计算机程序产品具体体现为软件产品，例如软件开发包(Software Development Kit，SDK)等等。
以上已经描述了本公开的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术的改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。

Claims (16)

  1. 一种对象检测方法，其特征在于，所述方法包括：
    确定至少一个误检区域属性，所述误检区域属性用于表征误检测区域的位置和尺寸；
    对待检测图像进行对象检测，得到至少一个对象检测结果，其中，所述至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性；
    根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
  2. 根据权利要求1所述的方法，其特征在于，所述确定至少一个误检区域属性包括：
    响应于人机交互操作生成的误检测区域，确定至少一个误检测区域属性。
  3. 根据权利要求2所述的方法，其特征在于，所述响应于人机交互操作生成的误检测区域，通过以下方式实现：
    对所述待检测图像进行对象检测并显示所述对象检测的预检测结果，所述预检测结果中包括至少一个初步检测框；
    响应于接收到用户通过人机交互操作选中的至少一个初步检测框，确定所述初步检测框表征的区域为误检测区域。
  4. 根据权利要求2所述的方法，其特征在于，所述响应于人机交互操作生成的误检测区域，通过以下方式实现：
    在场景显示界面显示编辑控件；
    响应于所述编辑控件被触发，显示与所述待检测图像对应的待编辑图像；
    响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域。
  5. 根据权利要求4所述的方法，其特征在于，所述响应于所述待编辑图像中的至少一个区域被选中，确定被选中所述区域为误检测区域包括：
    确定所述待编辑图像对应的至少一个图像框；
    确定所述至少一个图像框内的区域为被选中的误检测区域。
  6. 根据权利要求5所述的方法，其特征在于，所述确定所述待编辑图像对应的至少一个图像框包括：
    响应于接收到至少一个图像框坐标，在所述待编辑图像上确定所述至少一个图像框坐标对应的图像框；或者
    响应于接收到对应于所述待编辑图像的至少一个预设手势，根据所述至少一个预设手势投影在所述待编辑图像上确定的至少一个区域作为图像框。
  7. 根据权利要求1-6中任意一项所述的方法，其特征在于，所述对象检测结果中包括检测框坐标和对应的概率值，所述概率值用于表征所述待检测图像在对应检测框坐标内存在目标对象的可能性，所述检测框坐标为通过坐标方式表征的检测区域属性。
  8. 根据权利要求7所述的方法，其特征在于，所述根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果包括：
    根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选；
    根据所述对象检测结果中的概率值，对所述对象检测结果进行第二次筛选；
    根据第一次筛选和第二次筛选后得到的对象检测结果确定入侵检测结果。
  9. 根据权利要求8所述的方法，其特征在于，所述根据所述误检区域属性与所述检测区域属性的对应关系，对所述对象检测结果进行第一次筛选包括；
    响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，删除所述对象检测结果中的所述对象对应的检测区域属性。
  10. 根据权利要求9所述的方法，其特征在于，所述响应于所述对象检测结果中的对象位于所述误检区域属性表征的误检测区域内部，通过以下方式实现：
    确定所述对象检测结果中的对象的关键点位置；
    响应于所述关键点位置在一个误检区域属性表征的误检区域内部，确定所述检测区域属性表征得到检测区域在所述误检区域内部。
  11. 根据权利要求8-10中任意一项所述的方法，其特征在于，所述根据所述对象检测结果中的概率值进行第二次筛选包括：
    响应于对象检测结果中的概率值小于预设阈值，删除所述对象检测结果。
  12. 根据权利要求1-11中任意一项所述的方法，其特征在于，所述方法还包括：
    通过所述入侵检测结果标注所述待检测图像；
    输出标注后的待检测图像和对应的提示信息。
  13. 一种对象检测装置，其特征在于，所述装置包括：
    属性确定模块，用于确定至少一个误检区域属性，所述误检区域属性用于表征误检测区域的位置和尺寸；
    图像检测模块，用于对待检测图像进行对象检测，得到至少一个对象检测结果，其中，所述至少一个对象检测结果包括表征检测对象的位置和尺寸的检测区域属性；
    结果筛选模块，用于根据所述误检区域属性和所述检测区域属性对所述对象检测结果进行筛选，得到入侵检测结果。
  14. 一种电子设备，其特征在于，包括：
    处理器；
    用于存储处理器可执行指令的存储器；
    其中，所述处理器被配置为调用所述存储器存储的指令，以执行权利要求1至12中任意一项所述的方法。
  15. 一种计算机可读存储介质，其上存储有计算机程序指令，其特征在于，所述计算机程序指令被处理器执行时实现权利要求1至12中任意一项所述的方法。
  16. 一种计算机程序产品，包括计算机可读代码，或者承载有计算机可读代码的非易失性计算机可读存储介质，其特征在于，当所述计算机可读代码在电子设备的处理器中运行时，所述电子设备中的处理器执行用于实现权利要求1-12中的任意一项所述的方法。
PCT/CN2022/071050 2021-09-27 2022-01-10 对象检测方法及装置、电子设备和存储介质 Ceased WO2023045185A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111135692.7A CN113869295B (zh) 2021-09-27 2021-09-27 对象检测方法及装置、电子设备和存储介质
CN202111135692.7 2021-09-27

Publications (1)

Publication Number Publication Date
WO2023045185A1 true WO2023045185A1 (zh) 2023-03-30

Family

ID=78991101

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/071050 Ceased WO2023045185A1 (zh) 2021-09-27 2022-01-10 对象检测方法及装置、电子设备和存储介质

Country Status (2)

Country Link
CN (1) CN113869295B (zh)
WO (1) WO2023045185A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113869295B (zh) * 2021-09-27 2025-09-16 上海商汤智能科技有限公司 对象检测方法及装置、电子设备和存储介质
CN116245857A (zh) * 2023-03-07 2023-06-09 深圳绿米联创科技有限公司 目标检测方法、装置、电子设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090180670A1 (en) * 2006-05-09 2009-07-16 Hiroshi Iwamura Blocker image identification apparatus and method
CN109859177A (zh) * 2019-01-17 2019-06-07 航天新长征大道科技有限公司 基于深度学习的工业射线影像评定方法及装置
CN112115913A (zh) * 2020-09-28 2020-12-22 杭州海康威视数字技术股份有限公司 图像处理方法、装置及设备、存储介质
CN113869295A (zh) * 2021-09-27 2021-12-31 上海商汤智能科技有限公司 对象检测方法及装置、电子设备和存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007325170A (ja) * 2006-06-05 2007-12-13 Oki Electric Ind Co Ltd 映像監視装置及び方法
CN102411703A (zh) * 2010-09-21 2012-04-11 索尼公司 检测图像序列中的特定对象的装置、方法和视频摄像设备
KR102348593B1 (ko) * 2017-10-26 2022-01-06 삼성에스디에스 주식회사 기계 학습 기반의 객체 검출 방법 및 그 장치
CN110781821B (zh) * 2019-10-25 2022-11-01 上海商汤智能科技有限公司 基于无人机的目标检测方法及装置、电子设备和存储介质
CN112700657B (zh) * 2020-12-21 2023-04-28 阿波罗智联(北京)科技有限公司 用于生成检测信息的方法、装置、路侧设备和云控平台
CN112926510A (zh) * 2021-03-25 2021-06-08 深圳市商汤科技有限公司 异常驾驶行为识别方法及装置、电子设备和存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090180670A1 (en) * 2006-05-09 2009-07-16 Hiroshi Iwamura Blocker image identification apparatus and method
CN109859177A (zh) * 2019-01-17 2019-06-07 航天新长征大道科技有限公司 基于深度学习的工业射线影像评定方法及装置
CN112115913A (zh) * 2020-09-28 2020-12-22 杭州海康威视数字技术股份有限公司 图像处理方法、装置及设备、存储介质
CN113869295A (zh) * 2021-09-27 2021-12-31 上海商汤智能科技有限公司 对象检测方法及装置、电子设备和存储介质

Also Published As

Publication number Publication date
CN113869295A (zh) 2021-12-31
CN113869295B (zh) 2025-09-16

Similar Documents

Publication Publication Date Title
US11288531B2 (en) Image processing method and apparatus, electronic device, and storage medium
CN106970754B (zh) 截屏处理的方法及装置
CN114078118A (zh) 缺陷检测方法及装置、电子设备和存储介质
CN110287874A (zh) 目标追踪方法及装置、电子设备和存储介质
EP3301557A1 (en) Method, apparatus and computer program product for sharing content
CN110647834A (zh) 人脸和人手关联检测方法及装置、电子设备和存储介质
WO2022134382A1 (zh) 图像分割方法及装置、电子设备和存储介质、计算机程序
CN108509232A (zh) 屏幕录制方法、装置及计算机可读存储介质
EP3147802B1 (en) Method and apparatus for processing information
WO2016192325A1 (zh) 视频文件的标识处理方法及装置
WO2022134390A1 (zh) 标注方法及装置、电子设备和存储介质
CN110990801A (zh) 信息校验方法及装置、电子设备和存储介质
WO2023040202A1 (zh) 人脸识别方法及装置、电子设备和存储介质
CN106354504B (zh) 消息显示方法及装置
CN108829473A (zh) 事件响应方法、装置及存储介质
CN106331328B (zh) 信息提示的方法及装置
CN105353938A (zh) 悬浮短信显示方法及装置
CN110807393A (zh) 基于视频分析的预警方法及装置、电子设备和存储介质
CN113869295B (zh) 对象检测方法及装置、电子设备和存储介质
CN106650569B (zh) 指纹录入方法及装置
CN112381223A (zh) 神经网络训练与图像处理方法及装置
CN111832455A (zh) 获取内容图像的方法、装置、存储介质和电子设备
CN107580142B (zh) 一种执行处理的方法和装置
CN107977127B (zh) 更新页面的方法、装置及终端
CN112083841B (zh) 信息输入方法、装置和存储介质

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22871253

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 22871253

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 03/12/2024)

122 Ep: pct application non-entry in european phase

Ref document number: 22871253

Country of ref document: EP

Kind code of ref document: A1