Object position detection method and apparatus, storage medium, and electronic device

Abstract

Embodiments of the present disclosure provide an object position detection method and device, a storage medium and an electronic device, wherein the method can comprise: obtaining a target image, the target image comprising a first object and a second object; performing object detection processing on the target image to identify a reference detection box corresponding to a reference region of the first object and an object detection box corresponding to the second object; performing region expansion based on the reference detection box to obtain an expanded detection box, and determining a contact position of the second object on a dangerous surface based on the object detection box; and in response to the contact position being within the expanded detection box, determining that the second object has entered a dangerous region corresponding to the first object. Embodiments of the present disclosure not only save manpower and resources, but also improve the efficiency and accuracy of recognition.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to computer vision technology, and in particular, to an object position detection method and device, a storage medium and an electronic device. BACKGROUND

[0002] Safety accidents can occur in various scenarios, such as factories, construction sites, gas stations, etc. Taking a construction site as an example, when excavators are working at the construction site, the buckets of the excavators sometimes rotate, and if there are workers within a certain range of the excavator, the bucket will hit the workers, which can be life-threatening. The occurrence of safety accidents not only causes harm to the physical safety of workers, but also can cause the factory or construction site to stop working, which can have a significant impact on the interests of the enterprise. Therefore, it is necessary to detect whether there are people in the area around some equipment that poses a safety risk in order to prevent the occurrence of safety accidents. SUMMARY

[0003] The embodiments of the present disclosure at least provide an object position detection method and device, a storage medium and an electronic device.

[0004] In one aspect, an object position detection method is provided, the method comprising:

[0005] obtaining a target image, the target image comprising a first object and a second object;

[0006] performing object detection processing on the target image to identify a reference detection box corresponding to a reference region of the first object and an object detection box corresponding to the second object; the reference region is used to represent a contact region of the first object with a dangerous surface;

[0007] performing region expansion based on the reference detection box to obtain an expanded detection box, the expanded detection box being used to define a dangerous region of the first object on the dangerous surface; and determining a contact position of the second object on the dangerous surface based on the object detection box;

[0008] in response to the contact position being within the expanded detection box, determining that the second object enters the dangerous region corresponding to the first object.

[0009] In some embodiments, the first object is a vehicle; and the identifying the reference detection box corresponding to the reference region of the first object comprises: identifying a vehicle component included in the vehicle, the vehicle component being in contact with the dangerous surface; and determining the reference detection box corresponding to the reference region based on the vehicle component.

[0010] In some embodiments, the target image comprises a ground surface, and the vehicle is located on the ground surface. The vehicle component is a component of the vehicle in contact with the ground surface, and the dangerous surface is the ground surface,

[0011] In some embodiments, the region expansion based on the reference detection box to obtain an expanded detection box, the expanded detection box being used to define a dangerous region of the first object on the dangerous surface, comprises: obtaining region setting information of a dangerous region corresponding to the object category according to the object category of the first object; and performing region expansion on the reference detection box based on the region setting information to obtain an expanded detection box.

[0012] In some embodiments, the determination of the contact position of the second object on the dangerous surface based on the object detection box comprises: in a case where the dangerous surface corresponds to a lower edge of the object detection box, determining the contact position from the lower edge of the object detection box.

[0013] In some embodiments, the method further comprises: after determining that the second object enters the dangerous region corresponding to the first object, performing a region intrusion alarm.

[0014] In another aspect, an object position detection device is provided, the device comprising:

[0015] An image acquisition module, configured to acquire a target image, the target image comprising a first object and a second object;

[0016] An object recognition module, configured to perform object detection processing on the target image, and recognize a reference detection box corresponding to a reference region of the first object and an object detection box corresponding to the second object; the reference region being used to represent a contact region of the first object and a dangerous surface;

[0017] An expansion processing module, configured to perform region expansion based on the reference detection box to obtain an expanded detection box, the expanded detection box being used to define a dangerous region of the first object on the dangerous surface; and determine a contact position of the second object on the dangerous surface based on the object detection box;

[0018] A comparison processing module, configured to, in response to the contact position being within the expanded detection box, determine that the second object enters the dangerous region corresponding to the first object.

[0019] In some embodiments, the object recognition module, when used to recognize the reference detection box corresponding to the reference region of the first object, comprises: recognizing a vehicle component included in the vehicle, the vehicle component being in contact with the dangerous surface; and determining the reference detection box corresponding to the reference region according to the vehicle component; wherein the first object is a vehicle.

[0020] In some embodiments, the target image comprises a ground surface, and the vehicle is located on the ground surface; the vehicle component is a component of the vehicle in contact with the ground surface, and the dangerous surface is the ground surface.

[0021] In some embodiments, the expansion processing module, when used to expand the reference detection frame to obtain an expanded detection frame, comprises: obtaining region setting information of a dangerous region corresponding to the object category of the first object according to the object category of the first object; and expanding the reference detection frame to obtain an expanded detection frame based on the region setting information.

[0022] In some embodiments, the expansion processing module, when used to determine the contact position of the second object on the dangerous surface based on the object detection frame, comprises: in a case where the dangerous surface corresponds to a lower edge of the object detection frame, determining the contact position from the lower edge of the object detection frame.

[0023] In some embodiments, the device further comprises an alarm module configured to perform a region intrusion alarm after determining that the second object enters the dangerous region corresponding to the first object.

[0024] In another aspect, an electronic device is also provided, comprising a memory and a processor, wherein the memory is configured to store computer readable instructions, and the processor is configured to invoke the computer instructions to implement the method described in any of the embodiments of the present disclosure.

[0025] In another aspect, a computer readable storage medium is also provided, which stores a computer program, and the program is executed by a processor to implement the method described in any of the embodiments of the present disclosure.

[0026] The object position detection method and device, storage medium and electronic device provided by the embodiments of the present disclosure can determine whether the second object enters the dangerous region of the first object based on the same dangerous surface as much as possible, and further obtain the contact position of the second object on the dangerous surface according to the object detection frame, and then determine whether the contact position of the second object falls into the dangerous region based on the dangerous surface. This determination method can improve the accuracy of determining whether the second object enters the dangerous region. Therefore, the method of the embodiments can not only automatically monitor whether there is an object that is not allowed in the dangerous region, saving the consumption of manpower and material resources, but also improve the efficiency and accuracy of identification. BRIEF DESCRIPTION OF DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the one or more embodiments of the present disclosure or the related art, the drawings needed to be used in the embodiment or related art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments described in the one or more embodiments of the present disclosure, and those skilled in the art can obtain other drawings according to these drawings without any creative labor.

[0028] Figure 1A safety monitoring diagram for excavating a minefield is shown;

[0029] Figure 2 A flowchart of a method for detecting a position of an object is shown;

[0030] Figure 3 A dangerous area detection diagram is shown;

[0031] Figure 4 A flowchart of a method for detecting a position of an object is shown;

[0032] Figure 5 A structure diagram of an apparatus for detecting a position of an object is shown;

[0033] Figure 6 A structure diagram of an apparatus for detecting a position of an object is shown. DETAILED DESCRIPTION

[0034] In order to enable personnel in the art to better understand the technical solutions in the one or more embodiments of the present disclosure, the technical solutions in the one or more embodiments of the present disclosure will be described clearly and completely below in conjunction with the accompanying drawings of the one or more embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. Based on the one or more embodiments of the present disclosure, all other embodiments obtained by a person of ordinary skill in the art without creative labor should fall within the scope of protection of the present disclosure.

[0035] With the development of artificial intelligence technology, in the related art, AI technology is used for monitoring and intelligent analysis of safety accidents, to intelligently alarm dangerous behaviors in a scene such as a factory or a construction site, so as to avoid risks. This technology has largely liberated safety supervisors.

[0036] However, the accuracy of safety monitoring in the prior art is low. Please refer to Figure 1 As shown in FIG. 11, a safety supervisor of a construction site is taken as an example, an excavator 11 on the construction site is a device with safety risks, and workers are generally not allowed within a certain range around the excavator, so as to prevent being injured by a bucket of the excavator. Therefore, when safety monitoring is performed, it is necessary to monitor whether a worker has fallen into a dangerous area of the excavator 11, and if so, an alarm can be given in time to avoid a safety accident. In the related art, an image of the construction site can be collected, such as Figure 1The shown construction site image is subjected to object detection processing by a target detection network to obtain a vehicle body detection frame 12 corresponding to the vehicle body of the excavator 11, and a worker detection frame corresponding to a worker in the construction site image can also be detected. Then, the vehicle body detection frame 12 is expanded to obtain a danger area 13, and then it is judged whether the center point position of the worker detection frame falls into the danger area 13. By Figure 1 It can be seen that the center points of the worker detection frame 141 and the worker detection frame 142 both fall into the danger area 13.

[0037] However, the above related technology has a misjudgment situation. For example, the worker corresponding to the worker detection frame 141 is actually standing at a position far away from the excavator 11, so the person looks small, and is just framed into the frame of the danger area 13 because it is located above the excavator 11 in the construction site image. The worker detection frame 142 is the worker who is actually close to the excavator 11, and is also displayed high in the construction site image (the closer to the camera, the larger, and the farther away, the smaller). Therefore, it can be seen that the detection method in the related technology can cause false alarms, and the accuracy needs to be improved.

[0038] Based on this, Figure 2 A flowchart of a method for detecting the position of an object is shown, which is provided by at least one embodiment of the present disclosure, as shown in Figure 2 The method can include the following processing, as shown in

[0039] In step 200, a target image is obtained, and the target image includes a first object and a second object.

[0040] The target image can be an image collected in a monitoring scene. For example, if the monitoring scene is a construction site, it can be a construction site image collected by a camera in the construction site.

[0041] The first object and the second object can be objects to be monitored in the monitoring scene, and the first object and the second object can be associated objects with safety hazards. For example, in the case of a construction site, the excavator bucket may injure a worker, so the excavator and the worker are associated objects with safety hazards. In this embodiment, the first object can be the applicator of a safety accident, and the second object can be the applicator. For example, in the case of an excavator injuring a worker, the excavator is the stronger party and the applicator of the injury, and the worker is usually the injured party and the applicator, so the excavator can be referred to as the first object and the worker can be referred to as the second object.

[0042] The number of the first object and the second object in the target image can be at least one. For example, there can be multiple excavators in the image, and there can also be multiple workers. In the case where multiple first objects are included in the target image, such as multiple excavators are included in the construction site image, when the method of the embodiment is executed, for any excavator therein, it can be judged according to the method of the embodiment whether there is a worker entering the dangerous area of the excavator.

[0043] In step 202, object detection processing is performed on the target image to identify a reference detection box corresponding to a reference area of the first object and an object detection box corresponding to the second object; the reference area is used to represent the contact area of the first object and the dangerous surface.

[0044] In this step, the target image can be subjected to object recognition by a target detection network. The embodiment does not limit the specific network structure of the object detection network, which can include but is not limited to R-CNN, Fast-RCNN, R-FCN, etc.

[0045] The reference area can be understood as follows: taking an example of a working excavator on the ground of a construction site, the "10-meter range around the excavator" can be determined as a dangerous area where workers are prohibited to enter. In the definition of the dangerous area, the contact area of the excavator and the ground is taken as the reference, and the area is expanded outward by 10 meters. If it is a caterpillar excavator, the contact area of the caterpillar and the ground can be taken as the reference to expand outward by 10 meters to obtain the dangerous area.

[0046] In the embodiment, the reference area is used to represent the contact area of the first object and the dangerous surface. The first object is, for example, the excavator described above, and the dangerous surface is, for example, the ground described above. It can be understood as follows: the dangerous surface can be determined according to the setting of the dangerous area, such as if the "10-meter range around the excavator" is determined as the dangerous area, it is generally considered that no one is allowed to be within the 10-meter range around the contact area of the excavator and the ground, and therefore the ground can be considered as the dangerous surface defining the dangerous area. In specific implementations, the dangerous surface can be different according to different safety monitoring scenarios and settings of the dangerous area. For example, if the safety monitoring scenario is a slope, and the dangerous area is within a certain distance around the dangerous vehicle on the slope, the dangerous surface can also be the slope; if the safety monitoring scenario is a water surface, the dangerous surface can also be the water surface.

[0047] The embodiment does not limit the specific way of identifying the reference detection box corresponding to the reference area of the first object. For example, the vehicle components included in the vehicle can be identified by object detection processing, the vehicle components are in contact with the dangerous surface, and then the reference detection box corresponding to the reference area is determined according to the vehicle components. Still taking the example of the excavator, the contact area of the caterpillar and the ground can be taken as the reference to expand outward by 10 meters to obtain the dangerous area. Figure 1For example, in the case of the excavator, the track of the crawler excavator is in contact with the ground, so the track can be identified, i.e., the identified vehicle component is the track, and the detection box of the track is taken as the reference detection box corresponding to the reference region.

[0048] It should be noted that although the detection box of the track is not necessarily the strictly accurate contact area of the excavator with the ground, from the target image, the detection box of the track can basically represent the contact area of the excavator with the ground, i.e., the detection box is closest to the contact area of the excavator with the ground, so the detection box of the track can be taken as the reference detection box. Similarly, assuming that the first object is a large truck, when detecting the reference detection box, the wheels of the large truck can be detected first (i.e., the detected vehicle component is the wheel), and the area surrounded by the wheels at the edge of the large truck is taken as the reference region, which can represent the contact area of the large truck with the ground, so as to obtain the reference detection box corresponding to the reference region. Therefore, the detection method of the reference detection box can be determined according to the characteristics of different first objects and the situation of the danger area.

[0049] The object detection box corresponding to the second object can be at least a part of the second object, as long as the contact position of the second object on the danger area can be determined according to the object detection box in the subsequent step. For example, taking the second object as a human body, the object detection box of the second object can be the detection box of the whole human body, or can be the detection box of a part of the human body (for example, the lower body including the legs and feet of the human body), as long as the contact position of the feet with the ground can be determined according to the object detection box.

[0050] In addition, the second object in the embodiment is not limited to a human body, but can also be other objects such as cats, dogs and other animals. Any object that is not expected to be injured in the danger area of the first object can be used.

[0051] In step 204, the region expansion is performed based on the reference detection box to obtain an expanded detection box, and the expanded detection box is used to define the danger area of the first object on the danger area; and the contact position of the second object on the danger area is determined based on the object detection box.

[0052] The region expansion method can be determined according to the range setting of the danger area. For example, assuming that the "9-meter range around the excavator" is determined as the danger area, and the actual width of the track of the excavator is 6 meters, when the region expansion is performed, the track detection box of the excavator can be expanded by half of the width of the track. From the pixels of the target image, the width of the expanded region accounts for half of the width of the track. Please refer to the example of Figure 3 Figure 3 ​The crawler detection frame 31 of the excavator is recognized as a reference detection frame. And region expansion can be performed based on the crawler detection frame 31, for example, the crawler width can be expanded by half to obtain the danger region 32.

[0053] In addition, as Figure 3 shown, taking the second object as a human body as an example, the embodiment determines, based on the object detection frame 33, that the contact position of the second object on the danger surface is the contact position 34 of the human body with the ground.

[0054] In step 206, in response to the contact position being in the expanded detection frame, it is determined that the second object enters the danger region corresponding to the first object.

[0055] In this step, if the contact position of the second object on the danger surface determined based on the object detection frame is located in the expanded detection frame, it can be determined that the second object enters the danger region of the first object.

[0056] In the embodiment, when judging whether the second object enters the danger region of the first object, it can be judged as much as possible based on the same danger surface, for example, the reference detection frame is obtained based on the reference region of the contact between the first object and the danger surface, and the contact position of the second object with the danger surface is also obtained according to the object detection frame, then whether the contact position of the second object falls into the danger region is judged in the danger surface. This judgment method can improve the accuracy of the judgment of whether the second object enters the danger region. Therefore, the method of the embodiment not only can automatically monitor whether there is an object not allowed to exist in the danger region, saves the consumption of manpower and material resources, but also can improve the efficiency and accuracy of identification.

[0057] Taking the judgment of whether the worker enters the danger region of the excavator in the construction site image as an example, the danger region of the excavator is obtained by expanding the crawler detection frame, and the crawler detection frame is the representation as much as possible close to the actual contact area between the excavator and the ground in the construction site image. Figure 3 The danger region 32 obtained by expanding the crawler detection frame 31 is more accurate than the danger region 13 obtained by expanding the vehicle detection frame 12 of the vehicle body of the excavator in Figure 1 Since the crawler detection frame 31 is closer to the contact area between the excavator and the ground than the vehicle detection frame 12, the danger region 32 is more accurate than the danger region 13. Similarly, Figure 1 The center point of the worker detection frame 141 or the worker detection frame 142 in Figure 3The contact position 34 in the target image is closer to the contact position of the human body with the ground than the center point described above. Therefore, the contact position 34 is the contact position of the human body with the ground, and the dangerous area 32 is closer to the contact area of the excavator with the ground. Whether the contact position 34 falls into the dangerous area 32 is determined on the ground (dangerous surface), and the determination result is more accurate. By comparison, in the prior art, the dangerous area 13 is obtained by expanding the vehicle body detection frame 12, the vehicle body detection frame 12 includes the entire excavator, and the dangerous area obtained by expanding the vehicle body detection frame 12 is inaccurate. In addition, the vehicle body detection frame 12 is only a representation of the human body and cannot represent the contact position of the human body with the ground, so the comparison result is prone to misjudgment.

[0058] It should be noted that the ground in the above example is taken as an example of the dangerous surface, and the ground here is not limited to a flat surface. For example, the ground where the excavator is located is slightly lower than the ground where the worker stands, but they are both on the ground. Of course, the present embodiment does not limit other possible determination methods. For example, if the ground where the worker stands is higher than the ground where the excavator is located, the position conversion can be performed according to the object detection frame to obtain the contact position of the worker when the worker's position is converted to the ground where the excavator is actually located. For example, the actual contact position of the worker with the ground determined based on the object detection frame can be lowered by a certain height, which can be lowered to the ground plane where the excavator is located, to obtain the contact position of the worker on the dangerous surface for subsequent comparison. Whether the above conversion is performed or not, it is equivalent to comparing the contact position of the second object with the position of the expanded detection frame in the same dangerous surface.

[0059] The object position detection method of the present embodiment will be described below with reference to a specific scenario. Figure 4 A flowchart of an object position detection method provided by at least one embodiment of the present disclosure is shown, which takes monitoring whether a worker enters a dangerous area of a dangerous device on a construction site as an example. The method can include:

[0060] In step 400, a target image is obtained, and the target image includes a worker and an excavator.

[0061] For example, a camera can be installed on a construction site to capture a construction site image. For example, a construction site image as shown in FIG. 1 can be captured. Figure 1 The construction site image shown in FIG. 1 is a target image. The construction site image can include a plurality of workers and a construction device under construction, which includes but is not limited to an excavator and the like.

[0062] Wherein, there can be some potentially dangerous equipment in the construction equipment, and the embodiment takes the excavator as an example, in order to prevent the bucket of the excavator from hurting the workers, a certain range around the excavator can be defined as a dangerous area, and workers are not allowed to exist in the dangerous area.

[0063] The execution equipment of the method of the embodiment can be the camera described above, that is, the construction site image can be collected by the image collection equipment and the image can be processed; or the collected construction site image can also be transmitted to the backend processing equipment by the camera, and the method of the embodiment is executed on the backend processing equipment.

[0064] In step 402, object detection processing is performed on the target image to identify the track included in the excavator, and a track detection box is obtained, which is a reference detection box corresponding to the reference area of the excavator. The object detection box corresponding to the worker in the target image is also identified.

[0065] Target detection can be performed in this step to identify the track in the construction site image, which can be combined with reference to the schematic of Figure 3 The track detection box 31 can be detected. Each worker in the construction site image can also be detected to obtain an object detection box 33, and each worker can correspond to an object detection box 33. Wherein, the contact area between the excavator and the ground can be referred to as a reference area, and the track detection box 31 is a reference detection box in the construction site image that is as close as possible to the reference area.

[0066] In step 404, an extended detection box is obtained by region expansion from the track detection box, and the extended detection box is used to define the dangerous area of the excavator on the ground; and the contact position of the worker with the ground is determined based on the object detection box.

[0067] The definition of the corresponding dangerous area of different dangerous equipment can be different. For example, the dangerous area of the excavator can be defined as a range of 9 meters around, and the dangerous area of another large vehicle can be defined as a range of 5 meters around the large vehicle. That is, different dangerous equipment can use different settings of the dangerous area, and the setting of the dangerous area can also be determined according to the actual scene requirements.

[0068] The method execution device of the embodiment can pre-store the mapping relationship between objects of different categories and their dangerous areas, and when an object is detected by recognizing an image, the corresponding dangerous area setting information of the object category is found, and then the area is expanded. For example, the object category of the first object obtained by target detection on the construction site image is a excavator, and the area setting information of the dangerous area corresponding to the excavator is obtained, such as "a range of 9 meters around", and then the extended detection frame is obtained based on the track detection frame obtained in step 402. The extended detection frame represents the dangerous area of the excavator on the ground.

[0069] In addition, in an example, the lower edge of the object detection frame corresponding to the worker can be used as the contact surface between the worker and the ground, so that the contact position between the worker and the ground can be determined by the lower edge. For example, the midpoint of the lower edge of the object detection frame can be used as the contact position. In other examples, the midpoint of the lower edge of the object detection frame is not limited, such as other points on the lower edge instead of the midpoint, or a point inside the object detection frame close to the lower edge, etc.

[0070] For example, if the worker's body is standing straight, it will be more accurate to use the midpoint of the lower edge of the object detection frame as the contact position between the worker and the ground. For example, if the worker's body is bent, it is not the midpoint of the object detection frame, but the point offset to the side of the midpoint that better represents the contact position between the worker and the ground. Therefore, the contact position between the worker and the ground can be determined according to the actual scene.

[0071] The embodiment does not limit the above-mentioned extended detection frame, or the way of determining the contact position between the worker and the ground based on the object detection frame. For example, based on the detection of the object detection frame, the midpoint of the object detection frame can be directly determined as the contact position between the worker and the ground. For example, the contact position between the worker and the ground can also be output by a trained neural network, the input of the neural network can be an image including the worker and the ground, and the output is the contact position between the worker and the ground, or the input of the neural network can be an object detection frame including the worker and an image of the ground, and the output is the contact position between the worker and the ground, etc.

[0072] In step 406, in response to the contact position being in the extended detection frame, it is determined that the second object enters the dangerous area corresponding to the first object, and a region intrusion alarm is performed.

[0073] After detecting that the worker has entered the dangerous area of the excavator, a region intrusion alarm can be performed. For example, the worker entering the dangerous area can be highlighted on the display screen of the back-end processing device, such as displaying an alarm mark, and can also be assisted by a sound alarm prompt to prompt that a worker has entered the dangerous area.

[0074] In this embodiment, when judging whether the second object has entered the dangerous area of the first object, the judgment can be based on the same dangerous surface as much as possible. This judgment method can improve the accuracy of the judgment of whether the second object has entered the dangerous area. Therefore, the method of this embodiment can not only automatically monitor whether there is an unallowed object in the dangerous area, saving manpower and material resources, but also improve the efficiency and accuracy of identification. For example, in the prior art, workers far from the camera can be misjudged to have entered the dangerous area, but after using the method of this embodiment, because the determination of the dangerous area is more accurate, and the contact position of the worker with the ground is also more accurate, it is finally determined in the Figure 1 that the worker far from the camera has not entered the dangerous area of the excavator, thereby improving the accuracy of position detection. Figure 3

[0075] To implement the object position detection method of the above-mentioned embodiments of the present disclosure, the embodiments of the present disclosure further provide an object position detection device. Figure 5 The structure of an object position detection device provided by at least one embodiment of the present disclosure is shown in a structural schematic diagram, and the structure of the device is briefly described as follows, and the specific processing of each module can be combined with the description of the method embodiments. As shown in Figure 5 the device can include an image acquisition module 51, an object identification module 52, an expansion processing module 53, and a comparison processing module 54.

[0076] In one example, the object identification module 52, when identifying the reference detection frame corresponding to the reference area of the first object, includes: identifying a vehicle component included in the vehicle, the vehicle component being in contact with the dangerous surface; determining the reference detection frame corresponding to the reference area according to the vehicle component; and wherein the first object is a vehicle.

[0077] In one example, the target image includes a ground surface, and the vehicle is located on the ground surface; the vehicle component is a component of the vehicle in contact with the ground surface, and the dangerous surface is the ground surface.

[0078] In one example, the expansion processing module 53, when performing region expansion based on the reference detection frame to obtain an expanded detection frame, includes: acquiring region setting information of a dangerous area corresponding to an object category of the first object according to the object category; and performing region expansion on the reference detection frame based on the region setting information to obtain an expanded detection frame.​

[0079] In one example, the extension processing module 53, when determining the contact position of the second object on the danger surface based on the object detection frame, comprises: determining the contact position by the lower edge of the object detection frame in the case that the danger surface corresponds to the lower edge of the object detection frame.

[0080] In one example, as shown in Figure 6 The apparatus further comprises an alarm module 55, configured to perform area intrusion alarm after determining that the second object enters the danger area corresponding to the first object.

[0081] Those skilled in the art should understand that one or more embodiments of the present disclosure can be provided as a method, a system or a computer program product. Therefore, one or more embodiments of the present disclosure can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, one or more embodiments of the present disclosure can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0082] The embodiments of the present disclosure further provide a computer-readable storage medium, which can store a computer program. The program is executed by a processor to implement the object position detection method described in any of the embodiments of the present disclosure.

[0083] The embodiments of the present disclosure further provide an electronic device, which comprises a memory and a processor. The memory is configured to store computer readable instructions. The processor is configured to invoke the computer instructions to implement the object position detection method described in any of the embodiments of the present disclosure.

[0084] In the embodiments of the present disclosure, "and / or" means at least one of the two, for example, "A and / or B" includes three schemes: A, B, and "A and B".

[0085] Each of the embodiments in the present disclosure is described in a progressive manner, and the same or similar parts of each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, the data processing device embodiment is basically similar to the method embodiment, so the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment.

[0086] The above described specific embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the acts or steps recited in the claims can be performed in a different order and still accomplish the desired results. Also, the process depicted in the figures does not necessarily require the particular order shown, or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0087] Embodiments of the subject matter and the functional operations described in this disclosure can be implemented in digital electronic circuitry, in tangibly-embodied computer software or firmware, in computer hardware, including the structures disclosed in this disclosure and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this disclosure can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible non-transitory program carrier for execution by, or to control the operation of, data processing apparatus. Alternatively or in addition, the program instructions can be encoded on an artificially generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.

[0088] The processes and logic flows described in this disclosure can be performed by one or more programmable computers executing one or more computer programs to perform the functions described in this disclosure by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit), and the apparatus can be implemented as special purpose logic circuitry.

[0089] Suitable computers for executing computer programs include, for example, general-purpose and / or special-purpose microprocessors, or any other type of central processing unit. Typically, the central processing unit receives instructions and data from read-only memory and / or random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices for storing data, such as disks, magneto-optical disks, or optical disks, or the computer will be operatively coupled to such mass storage devices to receive data from or transfer data to them, or both. However, a computer is not required to have such devices. Furthermore, a computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.

[0090] Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. Processors and memory may be supplemented by or incorporated into dedicated logic circuitry.

[0091] While this disclosure contains numerous specific implementation details, these should not be construed as limiting the scope of any disclosure or the scope of the claims, but rather are primarily intended to describe the features of specific embodiments of a particular disclosure. Certain features described in the multiple embodiments of this disclosure may also be implemented in combination in a single embodiment. On the other hand, various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation of a sub-combination.

[0092] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0093] Thus, particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, actions recited in the claims can be performed in a different order and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0094] The above descriptions are only the preferred embodiment of one or more embodiments of the disclosure, and are not intended to limit one or more embodiments of the disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of the disclosure shall be included in the protection scope of one or more embodiments of the disclosure.

Claims

1. A method for detecting the location of an object, characterized in that, The method includes: Acquire a target image, wherein the target image includes a first object and a second object; The target image is subjected to object detection processing to identify the reference detection box corresponding to the reference region of the first object and the object detection box corresponding to the second object; the reference region is used to characterize the contact area between the first object and the dangerous surface; An extended detection frame is obtained by expanding the region based on the reference detection frame. The extended detection frame is used to define the dangerous area of ​​the first object on the dangerous surface. When the dangerous surface corresponds to the lower edge of the object detection frame, the contact position of the second object on the dangerous surface is determined by the lower edge of the object detection frame. In response to the contact location being within the extended detection frame, it is determined that the second object has entered the danger zone corresponding to the first object.

2. The method according to claim 1, characterized in that, The first object is a vehicle; The reference detection box corresponding to the reference region for identifying the first object includes: Identify vehicle components included in the vehicle that are in contact with the hazardous surface; Based on the vehicle components, a reference detection frame corresponding to the reference region is determined.

3. The method according to claim 2, characterized in that, The target image includes the ground, and the vehicle is located on the ground; The vehicle component is the component that contacts the ground, and the hazardous surface is the ground.

4. The method according to claim 1, characterized in that, The expanded detection frame is obtained by expanding the region based on the baseline detection frame. The expanded detection frame is used to define the danger zone of the first object on the danger surface, including: Based on the object category of the first object, obtain the area setting information of the danger zone corresponding to the object category; Based on the region setting information, the baseline detection box is expanded to obtain an expanded detection box.

5. The method according to any one of claims 1 to 4, characterized in that, The method further includes: after determining that the second object has entered the dangerous area corresponding to the first object, issuing an area intrusion alarm.

6. An object position detection device, characterized in that, The device includes: An image acquisition module is used to acquire a target image, wherein the target image includes a first object and a second object; An object recognition module is used to perform object detection processing on the target image, and to identify a reference detection box corresponding to the reference region of the first object and an object detection box corresponding to the second object; the reference region is used to characterize the contact area between the first object and the dangerous surface; An extended processing module is used to expand the region based on the reference detection frame to obtain an extended detection frame, the extended detection frame being used to define the danger area of ​​the first object on the danger surface; and when the danger surface corresponds to the lower edge of the object detection frame, the contact position of the second object on the danger surface is determined by the lower edge of the object detection frame. A comparison processing module is used to determine, in response to the contact location being within the extended detection frame, that the second object has entered the danger zone corresponding to the first object.

7. The apparatus according to claim 6, characterized in that, The object recognition module, when used to identify the reference detection box corresponding to the reference area of ​​the first object, includes: identifying vehicle components included in the vehicle, the vehicle components being in contact with the hazardous surface; and determining the reference detection box corresponding to the reference area based on the vehicle components; wherein the first object is a vehicle.

8. The apparatus according to claim 7, characterized in that, The target image includes the ground, and the vehicle is located on the ground; The vehicle component is the component that contacts the ground, and the hazardous surface is the ground.

9. The apparatus according to claim 6, characterized in that, The extended processing module, when used to expand the region based on the baseline detection box to obtain an extended detection box, includes: obtaining the region setting information of the danger zone corresponding to the object category according to the object category of the first object; and expanding the region based on the region setting information to obtain the extended detection box.

10. The apparatus according to any one of claims 6 to 9, characterized in that, The device further includes an alarm module, used to issue an area intrusion alarm after determining that the second object has entered the dangerous area corresponding to the first object.

11. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is used to store computer-readable instructions, and the processor is used to invoke the computer-readable instructions to implement the method according to any one of claims 1 to 5.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method described in any one of claims 1 to 5.

Images