WEIGHING SYSTEM AND WEIGHING METHOD WITH OBJECT RECOGNITION

Abstract

A weighing method comprises the steps of: recognizing one or more objects to be detected on a first scale platform top (A) or within an object recognition area of the first scale platform top (A), and weighing the objects to be detected that are placed on a second scale platform top (B). A weighing system comprises at least two scales having scale platform tops utilizing the weighing method outlined above. The weighing method reduces the difficulty of algorithm recognition by increasing the degree to which the object on the weighting platform fits the algorithm, reduces the complexity of operation flow and the time required, and effectively increases the precision and accuracy of object recognition.

Claims

1. A method of weighing one or more objects, comprising the steps of: recognizing, on a weigh platform of a first scale or within an object recognition area of the weigh platform, the one or more objects; and weighing, on a weigh platform of a second scale, the one or more objects.

2. The weighing method of claim 1, comprising the further steps of: acquiring detection results of the one or more objects that are respectively placed on the weigh platform of the first scale or within the object recognition areas thereof; and using the detection results in the step of recognizing the one or more objects.

3. The weighing method of claim 1, wherein: at least one first object and at least one second object, different from the first object, are recognized sequentially in the recognizing step; and the first object and second object are weighed sequentially in the weighing step, either according to a sequence in which the objects are recognized according to a pre-set sequence; or the different objects are sequentially on each of the first scales; and sequentially weighing the different objects that are placed on all of the second scales; either according to a sequence in which the objects are detected or a sequence in which the first scale weigh platforms or the object recognition areas thereof are arranged.

4. The weighing method of claim 3, comprising the following steps of: sequentially recognizing the different objects that are placed on the weigh platform or object recognition area of the first scale; recording them into a work sequence; and sequentially weighing the different objects that are placed on the weigh platform of the second scale according to the work sequence.

5. The weighing method of claim 1, wherein the step of recognizing the object comprises the sub-steps of: taking a picture including an image of the weigh platform of the first scale, or taking a picture of the object recognition area of the first scale; and using the picture to recognize the object to in the picture.

6. The weighing method of claim 5, wherein the step of recognizing the object further comprises a sub-step of: weighing the one or more objects that are placed on the first scale.

7. The weighing method of claim 6, wherein the recognizing step comprises a sub-step of: sending the picture to a training model to recognize the object; or sending both the picture and a weight value of the object to the training model.

8. The weighing method of claim 7, wherein: the training model is configured to recognizes the object in the picture by using a picture feature comparison; or the training model is configured to recognize the object in the picture by using the picture feature comparison and whether a deviation between the weight value of the object and a pre-set standard weight value in the training model is within an error range.

9. The weighing method of claim 8, wherein the training model is constructed by the steps of: taking pictures of one or more objects placed within the object recognition area on the first scale in different angular directions; and sending the pictures to a recognition algorithm for constructing the pictures into a training model.

10. The weighing method of claim 9, wherein the recognition algorithm for constructing the training model is sent the pictures taken and at least one of the following: weight information of the object, light source information, shadow information.

11. The weighing method of claim 1, further comprising the steps of: outputting a weight value and the recognized information of the object; or inputting the weight value and the recognized information of the object into an order or a database, or outputting a counted number of the objects after counting the objects by means of the weight value of the object.

12. A weighing system, comprising at least one first scale and at least one second scale, each having a weigh platform, wherein the weighing system is configured to perform the method steps of claim 1.

13. The weighing system of claim 12, wherein: each first scale in the weighing system has a precision that is the same as or higher than a precision of each second scale in the weighing system.

14. The weighing system of claim 12, wherein: each first scale and each second scale is provided with an image recognition device; and each first scale is operatively configurable as a second scale, and vice versa.

15. (canceled)

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0041] FIG. 1 is a schematic view of a scale platform top of a platform scale in a weighing system according to an embodiment of the present invention.

[0042] FIG. 2 is a schematic view of a scale platform top of a platform scale in a weighing system according to another embodiment of the present invention.

DETAILED DESCRIPTION OF EMBODIMENTS

[0043] The present invention is further illustrated below by way of embodiments, but is not thus limited within the scope of the embodiments.

[0044] A weighing system of the present invention comprises two or more scales, wherein one or more of the scales are selected as the scale for object recognition, and the objects are recognized by means of image recognition. All the objects on scale platforms of the other scales are then weighed, and the weighing data and object recognition results can be used to further count the objects and perform other such functions that are relevant in the context of retailing and industrial weighing. In addition, the processing of formulation or order information can be realized by combining article information related to the object.

[0045] By way of the embodiments as described below, the implementations of the present invention are illustrated by way of examples.

[0046] A weighing system comprising two platform scales in an embodiment is shown in FIG. 1, wherein a field of view of a camera on one high-precision platform scale A covers the entire scale platform of the scale, and in this embodiment image recognition is performed only on an object placed on the platform scale A of the weighing system. The other platform scale B is a platform scale with a low precision and is only used for weighing objects.

[0047] In another embodiment, a partial area of the platform scale A is selected that is covered by the field of view of the camera and serves as an object recognition area.

[0048] When performing object recognition, either one object or multiple objects, for example, three or six objects, may be placed on the platform scale A. The camera takes a picture of the scale platform top of the platform scale A, and the weight weight_a of the object is weighed on the platform scale A and the weight data is saved.

[0049] The background algorithm recognizes the object(s) placed on the weighing platform top of the platform scale A by means of the picture, extracts image features and compares the extracted image features with a previously stored training model to establish a relationship. The background algorithm gives an object matching confidence, and then compares the weight information weight_a with weight information weight_a_s recorded in the model within a pre-set certain tolerance. The confidence and the weight tolerance are combined with the object recognition result to make a comprehensive determination.

[0050] Then, an unlimited number of objects can be placed on or dumped to the platform scale B. In this embodiment, the object recognition is performed only on the platform scale A, so that there is no requirement for the placement of objects on the platform scale B compared to the prior art, which is convenient for the operator to use and operate.

[0051] Thereafter, the weight weight_ab on the entire platform top is weighed, and the weight data weight_a and weight_ab is sent to the background algorithm for processing.

[0052] When the background algorithm determines that the object on the weighing platform top of the platform scale A is the desired object, the object information and the weight information weight_ab can be directly output or recorded. It is also possible to send the weight data weight_ab on the entire platform top to an order to record the total weight of the objects in the order, and to calculate the number of the objects of this type by the relationship between the total weight and the piece weight (e.g., weight_a_s), and the quantity information is also sent to the order for storage. It is also possible to count the objects. After the number reaches the number indicated in the order, the relevant information is inputted into the order, and the platform scale of the weighing system proceeds to the processing of objects of the next type or waits for the next operation.

[0053] In another embodiment, the platform scale B has the same configuration as the platform scale A, and is also a high-precision scale equipped with a camera for image recognition. In this embodiment, the weighing system still uses one of the platform scales for object recognition and the other for performing object weighing, but the functions of the platform scale A and the platform scale B in the weighing system are interchangeable.

[0054] In another embodiment, the weighing system further comprises three platform scales identical to the platform scale B. These four platform scales B may be platform scales with a low precision. Weighing by multiple platform scales B can improve the rate of weighting or counting. In addition, by increasing the number of the platform scales B, the rate of weighing or counting can be further increased. At the same time, in case of same target weighing capacity or quantity, the mode with multiple platform scales B has reduced requirements for the weighing capacity of the single platform scale, so that the configuration of the platform scale in the weighing system is more diverse.

[0055] In still another embodiment, while object recognition is performed on the platform scale A, the platform scale B weighs the object(s) placed or dumped thereon. In addition, the platform scale B may first weigh the object placed or dumped thereon. The object to be recognized is then placed on the platform scale A, and the object is recognised by means of image recognition.

[0056] In a variant example, after the platform scale A completes recognition, and before the platform scale B completes the weighing and counting of the object, another object can be placed on the platform scale A for recognition. The process also includes waiting for the platform scale B to complete the current weighing task, and then the platform scale B performs the weighing task for the next object.

[0057] In another variant example, the platform scale A sequentially recognizes the different objects placed on the platform and generates a weighing task queue. The platform scale sequentially completes the weighing tasks according to the weighing task queue, and records them as individual orders one time.

[0058] To further reduce the difficulty of feature extraction and comparison of the algorithm, in another embodiment, the training model is also established by utilizing the platform scale A. In the model establishment, at first, one object, or multiple objects such as seven or twelve objects, is/are placed on the weighing platform top of the platform scale A in a specified posture(s) and position (s). Exemplary posture(s) and position(s) of the object(s) such as a part with three standing faces, only one standing face may be selected, a part is placed in the center of the weighing platform top with its orientation being perpendicular to the position of the weighing platform top. A picture is then taken and the taken picture is sent to the recognition algorithm for model training. Finally, the model is obtained, which contains information such as the object image and the presented posture.

[0059] In another embodiment, while a picture of the object is taken, the object is also weighed, and the weight information weight_a_s of the object is obtained and sent to the recognition algorithm for model training.

[0060] In still another embodiment, the environment information such as the light source and the shadow during the entire process of taking the picture or before or when taking the picture of the object is also sent to the recognition algorithm for model training.

[0061] In the embodiment shown in FIG. 2, the weighing system comprises two high-precision platform scales A1 and A2, and two platform scales B. The platform scales B have the same configuration as the platform scale B of the embodiment shown in FIG. 1. The platform scales A1 and A2 have the same configuration as the platform scale A in the embodiment shown in FIG. 1.

[0062] In this embodiment, the two platform scales A1 and A2 are both used to recognize the carried object, and in this embodiment, the recognition results of the two platform scales are integrated, and the final object recognition result is thus obtained. When, in a variant example of the embodiment, more platform scales having the same configuration as the platform scale A1 or the platform scale A2 are added to improve the recognition result, some existing calculation models or neural networks can be used to further improve the correct ratio of recognition results.

[0063] The two platform scales B respectively weigh the objects carried thereon, increasing the rate of weighing or counting.

[0064] In another embodiment, the platform scales A1 and A2 respectively recognize different objects, and the two platform scales B only perform weighing or counting of one type of objects, and after completion, perform weighing or counting of another type of objects. For example, for the platform scale B, the weighing and counting of two different types of objects are performed according to the priority levels of the two types of objects set in advance, or the weighing and counting are performed according to the sequence in which the objects are recognized. In this embodiment, two platform scales are used to quickly switch the weighing or counting of different objects, e.g., when the object recognized on the platform scale A1 is weighed on the platform scale B, the platform scale A2 detects the next object to be recognized, and the platform scale B can immediately switch to the weighing of the next object when it completes the weighing of the previous object. By switching the platform scale B to weigh the different objects recognized by the platform scales A1 and A2, the weighing sequence can be determined for different orders.

[0065] Compared with the method of the embodiment of FIG. 1, the platform scale A1 corresponds to one platform scale B, and the platform scale A2 corresponds to one platform scale B, so that the weighing system of this embodiment has a higher efficiency.

[0066] In still another variant example, the platform scale A3 and the platform scale A4 configured in the same manner as the platform scale A1 are also provided, so that four different types of objects can be simultaneously processed. Therefore, based on the number of the objects that need to be processed, more platform scales A can be provided.

[0067] Although the specific implementations of the present invention are described above, a person skilled in the art should understand that these are only exemplary, and the scope of protection of the present invention is defined by the attached claims. A person skilled in the art can make various changes or modifications to these implementations without departing from the principle and spirit of the present invention, but all the changes or modifications fall within the scope of protection of the present invention.

REFERENCE SIGNS LIST

[0068] A, A1, A2 high-precision platform scale

[0069] B low-precision platform scale