CN109190696A - A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing - Google Patents
A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing Download PDFInfo
- Publication number
- CN109190696A CN109190696A CN201810989007.9A CN201810989007A CN109190696A CN 109190696 A CN109190696 A CN 109190696A CN 201810989007 A CN201810989007 A CN 201810989007A CN 109190696 A CN109190696 A CN 109190696A
- Authority
- CN
- China
- Prior art keywords
- image
- package
- sorted
- pixel
- logistics
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Abstract
The embodiment of the invention discloses a kind of logistics package classification method, device, equipment and computer readable storage mediums.Wherein, method includes carrying out edge detection to the package image to be sorted acquired under default shooting angle and shooting background, obtains the objective contour image of package image to be sorted;The sum of pixel in objective contour image is calculated, according to the relational expression of pixel and geometric element, the space hold value of package image to be sorted is calculated, is classified according to the value to package to be sorted.The space hold value that technical solution provided by the present application is wrapped up according to logistics is classified, the parking space of logistics sortation hubs can not only be efficiently used, greatly improve the conevying efficiency of subsequent package, it also can avoid since sampled images randomness is high, and acquired image background it is complex caused by ambient noise the phenomenon that being affected to the quality of image, the effective accuracy for improving package dimensions and calculating, and then promote the precision and classification effectiveness of package classification.
Description
Technical field
The present embodiments relate to logistics Sorting Technique field, more particularly to a kind of logistics package classification method, device,
Equipment and computer readable storage medium.
Background technique
With the widespread development of Internet technology, the operation flow electronization of more and more commercial activitys, e-commerce (letter
Claim electric business) it applies and gives birth to.Increasingly raising and e-commerce transaction based on the living standards of the people wide, convenience with covering scope
Etc. advantages, e-commerce transaction become trend of the times.
With the quick universalness of E-business applications, closely coupled to it is logistics Sorting Technique, logistics sorting
A most key ring is the classification wrapped up in technology, to a certain extent, the working efficiency and package point of logistics sortation hubs
The accuracy of class has vital influence to the development of e-commerce.
In the prior art, package classification is generally to be divided based on destination address (the harvest address on such as electronic order)
Class, such as can successively be divided according to same city-of the same province-same cell in same area-, that is to say, that logistics sortation hubs
The package of same destination address would generally be divided into the same area, the division benchmark of same destination address can be according to current logistics
Geographical location locating for sortation hubs determines, such as current logistics sortation hubs are provincial city, then being sorted for the first time
When, it is divided on the basis of same province or same municipality directly under the Central Government, i.e., same province is classified as a major class.
Biggish commodity (such as fitness equipment, cup, TV are differed due to containing object volume in the article of online sales
Machine etc.), correspondingly, each inclusion enclave accumulates in the larger situation of great disparity, and commodity volume is different, required for the storage that occupies it is empty
Between size and transportation safety require to will be different, such as staple commodity usually requires fork truck, gravity type shelf etc., and small
Part commodity, which can usually borrow, rotates horizontally automatic shelf.
But the package classification based on destination address leads to logistics sortation hubs not in view of the volume great disparity of each package
The parking space that logistics package can be efficiently used, so that space resources be caused to waste, and is unfavorable for subsequent transportation.
Summary of the invention
The purpose of the embodiment of the present invention is that providing a kind of logistics package classification method, device, equipment and computer-readable depositing
Storage media improves the accuracy and classification effectiveness of logistics package classification.
In order to solve the above technical problems, the embodiment of the present invention the following technical schemes are provided:
On the one hand the embodiment of the present invention provides a kind of logistics package classification method, comprising:
Obtain the package image to be sorted under default shooting angle and shooting background;
Edge detection is carried out to the package image to be sorted, obtains the objective contour figure of the package image to be sorted
Picture;
The sum of pixel in the objective contour image is calculated, according to the relationship of the pixel and geometric element that pre-establish
The space hold value of the package image to be sorted is calculated, using as the mark classified to the package to be sorted in formula
It is quasi-.
Optionally, the objective contour image includes the target segment of main view and the profile of top view, the calculating institute
The sum of pixel in objective contour image is stated, according to the relational expression of the pixel and geometric element that pre-establish, institute is calculated
Stating the space hold value to be sorted for wrapping up image includes:
The package image to be sorted is top view, calculates the sum of interior pixels point of the corresponding profile of the top view,
According to the relational expression of the pixel and area that pre-establish, the area of the package image to be sorted is calculated;
The package image to be sorted is main view, the sum of pixel of target segment is calculated, according to the picture pre-established
The height of the package image to be sorted is calculated in the relational expression of vegetarian refreshments and length;The target segment is the main view
Contour images the corresponding line segment of the maximum height on y direction;
The volume of the corresponding package of the package image to be sorted is calculated according to the area and the height.
Optionally, after the package image to be sorted obtained under default shooting angle and shooting background, further includes:
Image preprocessing is carried out to the package image to be sorted;
It is described to include: to the package image progress image preprocessing to be sorted
To the package to be sorted carry out image gray processing, image threshold, in image filtering any one or it is any
Combination.
Optionally, the relational expression of the basis pre-establishes pixel and geometric element includes:
Obtain the figure of the known dimensions for the predetermined number drawn in default background;
Record the default line segment length of each figure and area and the corresponding pixel of corresponding number of pixels and each figure
Number, as two groups of test datas;
Two groups of test datas are fitted respectively using least square method, obtain pixel and length relational expression, with
And the relational expression of pixel and area.
Optionally, the relational expression of the pixel and geometric element that pre-establish in the basis is calculated described wait divide
Class is wrapped up after the space hold value of image, further includes:
Obtain package classification thresholds;
Judge whether the space hold value is greater than the classification thresholds;
If so, by the package to be sorted labeled as big package;
If it is not, the package to be sorted is then labeled as small packet.
On the other hand the embodiment of the present invention provides a kind of logistics package sorter, comprising:
Image capture module, for obtaining the package image to be sorted under default shooting angle and shooting background;
Edge detection module obtains the package to be sorted for carrying out edge detection to the package image to be sorted
The objective contour image of image;
Pixel computing module, for calculating the sum of pixel in the objective contour image;
Package dimensions computing module is calculated for the relational expression according to the pixel and geometric element pre-established
The space hold value of the package image to be sorted, using as the standard classified to the package to be sorted.
Optionally, the package dimensions computing module includes:
Areal calculation submodule is top view for the package image to be sorted, calculates the corresponding wheel of the top view
Wide the sum of interior pixels point, according to the relational expression of the pixel and area that pre-establish, is calculated the package to be sorted
The area of image;
Height computational submodule, for the package image to be sorted for main view, calculate target segment pixel it
With according to the relational expression of the pixel and length that pre-establish, the height of the package image to be sorted is calculated;The mesh
Graticule section is the corresponding line segment of the maximum height on y direction of the contour images of the main view;
Volume computational submodule, for the package image pair to be sorted to be calculated according to the area and the height
The volume for the package answered.
It optionally, further include image pre-processing module, for carrying out image gray processing, figure to the package image to be sorted
As any one or any combination in thresholding, image filtering.
The embodiment of the invention also provides a kind of logistics to wrap up sorting device, including processor, and the processor is for holding
The step of classification method is wrapped up in logistics as described in preceding any one is realized when the computer program stored in line storage.
The embodiment of the present invention finally additionally provides a kind of computer readable storage medium, the computer readable storage medium
On be stored with logistics package sort program, logistics package sort program is realized as described in preceding any one when being executed by processor
The step of classification method, is wrapped up in logistics.
The embodiment of the invention provides a kind of logistics to wrap up classification method, to acquiring under default shooting angle and shooting background
Package image to be sorted carry out edge detection, obtain it is to be sorted package image objective contour image;Calculate objective contour figure
The sum of pixel, according to the relational expression of the pixel and geometric element that pre-establish, is calculated package image to be sorted as in
Space hold value, classified according to the value to package to be sorted.
The advantages of technical solution provided by the present application is, the classification wrapped up using logistics package dimensions size as logistics
Standard reduces the greatly different degree for being classified as of a sort package volume, so as to efficiently use the storage sky of logistics sortation hubs
Between, greatly improve the conevying efficiency of subsequent package;In addition, the package image shot using predetermined angle, and pass through package
Contour images the size of package is calculated, can to avoid due to sampled images randomness it is relatively high, and acquire figure
The phenomenon that ambient noise caused by the background of picture is complex is affected to the quality of image, can effectively improve package
The accuracy that size calculates, and then promote the precision and classification effectiveness of package classification.
In addition, the embodiment of the present invention provides corresponding realization device, equipment and meter also directed to logistics package classification method
Calculation machine readable storage medium storing program for executing, further such that the method has more practicability, described device, equipment and computer-readable storage
Medium has the advantages that corresponding.
Detailed description of the invention
It, below will be to embodiment or existing for the clearer technical solution for illustrating the embodiment of the present invention or the prior art
Attached drawing needed in technical description is briefly described, it should be apparent that, the accompanying drawings in the following description is only this hair
Bright some embodiments for those of ordinary skill in the art without creative efforts, can be with root
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram that classification method is wrapped up in a kind of logistics provided in an embodiment of the present invention;
Fig. 2 is the least square method fitting schematic diagram of line segment provided in an embodiment of the present invention and pixel;
Fig. 3 is the least square method fitting schematic diagram of area provided in an embodiment of the present invention and pixel;
Fig. 4 is Laplacian edge detection results schematic diagram provided in an embodiment of the present invention;
Fig. 5 is Canny edge detection results schematic diagram provided in an embodiment of the present invention;
Fig. 6 is the package image to be sorted after image binaryzation provided in an embodiment of the present invention;
Fig. 7 is effect image of Fig. 6 image provided in an embodiment of the present invention after mean filter;
Fig. 8 is effect image of Fig. 6 image provided in an embodiment of the present invention after gaussian filtering;
Fig. 9 is effect image of Fig. 6 image provided in an embodiment of the present invention after median filtering;
Figure 10 is the flow diagram that classification method is wrapped up in another logistics provided in an embodiment of the present invention;
Figure 11 is the height calibration result schematic diagram of the main view of illustrative example provided in an embodiment of the present invention;
Figure 12 is a kind of specific embodiment structure chart that sorter is wrapped up in logistics provided in an embodiment of the present invention;
Figure 13 is another specific embodiment structure chart that sorter is wrapped up in logistics provided in an embodiment of the present invention.
Specific embodiment
In order to enable those skilled in the art to better understand the solution of the present invention, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.Obviously, described embodiments are only a part of the embodiments of the present invention, rather than
Whole embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present invention.
The description and claims of this application and term " first ", " second ", " third " " in above-mentioned attached drawing
Four " etc. be for distinguishing different objects, rather than for describing specific sequence.Furthermore term " includes " and " having " and
Their any deformations, it is intended that cover and non-exclusive include.Such as contain a series of steps or units process, method,
System, product or equipment are not limited to listed step or unit, but may include the step of not listing or unit.
After describing the technical solution of the embodiment of the present invention, the various non-limiting realities of detailed description below the application
Apply mode.
Referring first to Fig. 1, Fig. 1 is the flow diagram that classification method is wrapped up in a kind of logistics provided in an embodiment of the present invention,
The embodiment of the present invention may include the following contents:
S101: the package image to be sorted under default shooting angle and shooting background is obtained.
Shooting angle can be that the angle of the shooting package image to be sorted pre-set can optionally be shot to be sorted
Wrap up the top view and main view of image.Certainly, the image under other shooting angle can be also chosen, this does not influence the application's
It realizes.
Shooting background can be the shooting background of the fixation pre-set, to avoid since different shooting backgrounds are to subsequent figure
As processing impacts.
For the package categorizing system structure of the entire logistics sortation hubs of simplification, fine micro-displacement mechanism is avoided to design, it can
It directlys adopt distortionless industrial camera and carries out Image Acquisition, and the bandpass filter built in the industrial camera, Ke Yiyou
Effect filters out the light of wavelength other than visible light, achievees the purpose that enhance red visible, facilitates carry out image procossing.
S102: carrying out edge detection to package image to be sorted, obtains the objective contour image of package image to be sorted.
Image Edge-Detection (or edge enhancing) can make image outline or edge become more prominent, in addition to edge
Image by can all be weakened after this processing, or even be completely removed, be a kind of area-of-interest processing method.
The edge detection for wrapping up image wraps up one indispensable content of classification as logistics, wraps up ruler to logistics is calculated
Very little accuracy and rapidity plays a role.Perimeter brightness change in the brightness and original image on enhanced boundary
Rate is directly proportional.
When carrying out edge detection to package image to be sorted, reference can be made to any achievable edge inspection in the prior art
It surveys or the algorithm of edge enhancing, the application does not do any restriction to this.About concrete implementation process, herein, repeat no more.
S103: the sum of pixel in objective contour image is calculated.
Objective contour image is the three-dimensional information for calculating the space hold value of package to be sorted, and objective contour image can
Include the exterior contour for calculating two-dimensional areas and the line segment for calculating package height to be sorted.Correspondingly, objective contour
The sum of pixel is the sum of all pixels point on the sum of all pixels point inside exterior contour and line segment in image.
S104: according to the relational expression of the pixel and geometric element that pre-establish, package image to be sorted is calculated
Space hold value, using as the standard classified to package to be sorted.
The type of space hold value can be determined according to pre-set classification standard, if classification standard is package
Size, then corresponding space hold value can be the volume of package to be sorted;If classification standard is the height of package, that
Corresponding space hold value can be the space occupied in the longitudinal direction of package to be sorted, such as the deposit for different height
Material package rest stand (including multilayer, every layer between fixed height), can be high by rest stand interlayer in order to make full use of each rest stand
Degree is used as classification standard, i.e., will wrap up package of the volume height no more than interfloor height and be placed on the same rest stand.
All pixels point inside objective contour is summed, it is total to obtain all pixels point in package image to be sorted
With.If the simple sum for knowing pixel in image, can not find out the real area of object, and between pixel and area
Existing is not only simple linear relationship, is very likely to that there are nonlinear relationships, but between pixel and area
Nonlinear relationship needs are verified first with the formula derived in advance, namely need to pre-establish pixel and geometric element
Relational expression, then calculated using the relational expression pre-established.
In technical solution provided in an embodiment of the present invention, the classification wrapped up using logistics package dimensions size as logistics
Standard reduces the greatly different degree for being classified as of a sort package volume, so as to efficiently use the storage sky of logistics sortation hubs
Between, greatly improve the conevying efficiency of subsequent package;In addition, the package image shot using predetermined angle, and pass through package
Contour images the size of package is calculated, can to avoid due to sampled images randomness it is relatively high, and acquire figure
The phenomenon that ambient noise caused by the background of picture is complex is affected to the quality of image, can effectively improve package
The accuracy that size calculates, and then promote the precision and classification effectiveness of package classification.
In a kind of specific embodiment, when calculating the space hold value of package image to be sorted, it may include:
Package image to be sorted is top view, the sum of the interior pixels point of the corresponding profile of top view is calculated, according to preparatory
The area of package image to be sorted is calculated in the pixel of foundation and the relational expression of area;
Package image to be sorted is main view, the sum of pixel of target segment is calculated, according to the pixel pre-established
With the relational expression of length, the height of package image to be sorted is calculated;Target segment is the contour images of main view vertical
The corresponding line segment of maximum height in axis direction;
The volume of package corresponding with package image to be sorted is highly calculated according to area.
Parameter calibration is to determine the relationship between the length of line segment and pixel under fixed background and distance.Parameter calibration
Accuracy determine measurement article size accuracy.Using background calibration method, can measure between line segment and pixel
Relationship.And then relationship between object area and pixel is acquired, the relationship between object volume and pixel may finally be acquired.
It is fitted using least square method algorithm, fits the relationship between line segment length and pixel, or asking
When taking the relationship between area and pixel, all it is the parameters in the linear relation obtained, brings parameters into formula
In, and it is applied to the measurement of practical line segment and area.But error is also existing, because shot when calibration
Picture is it is difficult to ensure that its background all claps, because of this error component, even if several groups of photographs of very careful shooting
Piece cannot guarantee that the pixel that every picture is included is consistent with whole even the whole bats of previous background are entered as possible,
This ratio that will result in the line segment length and pixel actually demarcated is not absolutely accurate, but this can be by specially surveying
Amount, and the mode of industry camera is fixed to realize, or this task is completed by scaling method described above.Separately
One is that will lead to the object of certain altitude due to the relationship between shooting distance and focal length can be figure the reason of leading to error
As becoming larger, and before, calibration is the figure by drawing in background to seek line segment and pixel respectively, between area and pixel
Relational expression.
Specifically, the relational expression of the pixel and geometric element that pre-establish can are as follows:
The figure of eight known dimensions is drawn under fixed background, respectively, calculates the corresponding pixel of each line segment length
It counts and makes a record, in the same way, observe eight groups of data of area and pixel, and record, it then can be in Pycharm
Two groups of data are fitted respectively using least square method.Specific observation method is to draw known length under specific background
With the figure of area, carry out image procossing after, the number of pixel between the maximum value and minimum value of the ordinate of marker graphic,
The number of pixels of the number of pixel and each graphics area between the maximum value and minimum value of abscissa utilizes in this approach
Eight pictures calculate the number of pixels and eight area number of corresponding pixels of eight groups of line segments.
Eight groups of data are measured according to the above method, program operation result shows pixel number between abscissa, ordinate
Between pixel number and figure inside included all pixels.To measure actual length and area.It lifts
For example, the data measured can be as follows shown in Tables 1 and 2:
1 line segment of table and pixel number observe table
Line segment length | 5 | 6 | 7 | 8 | 10 | 11 | 13 | 16 |
Number of pixels | 130 | 170 | 200 | 232 | 299 | 331 | 401 | 492 |
Image after being fitted using least square method according to the data in above table is please referred to shown in Fig. 2.According to
Relational expression known to the figure between line segment and pixel meets linear relationship, and the relational expression being arranged between line segment and pixel is y
=ax+b, wherein y is practical line segment length, and x is the number of pixel in line segment in image, and a and b are two location parameters.By
This visible relationship using between pixel and practical line segment length not only improves practicability, but also reduces the quantity of parameter.In reality
Border operate when, only need to measure two groups known to object-image displacement, system parameter can be calculated.
According to the above numerical value, can simultaneous linear equation in two unknowns group, parameter a=0.02995, b=1.18223 can be obtained, that
Relational expression between line segment and pixel is y=0.02995x+1.18223.
After determining parameter value, it can be obtained using the system parameter solved above according to the displacement relative to base position
Line segment length to object plane to be measured relative to benchmark line segment.
Area and the observation value of pixel are as shown in table 2:
2 area of table and pixel number observe table
Image after being fitted using least square method according to the data in above table is please referred to shown in Fig. 3.According to
Relational expression known to the figure between area and pixel meets linear relationship, and the relational expression between setting area and pixel is s
=a'n+b', wherein s is real area, and n is the number of pixel in area, and a' and b' are two location parameters.It can be seen that
Using the relationship between pixel and real area equally parameter a'=can be obtained by simultaneous linear equation in two unknowns group
0.00011296440782052016, b'=5.700882225384185, then the relational expression between line segment and pixel is
Are as follows:
S=0.00011296440782052016n+5.700882225384185.
After determining parameter value, it can be obtained using the system parameter solved above according to the displacement relative to base position
Area to object plane to be measured relative to reference area.
From the foregoing, it will be observed that the application is pre-established between pixel and line segment length by parameter calibration and least square method
Relational expression between relational expression and area and pixel can effectively improve accuracy and effect that package volume to be sorted calculates
Rate.
The algorithm of general common edge detection may include Roberts operand, Sobel (Sobel) operand, Prewitt
Operand, Kirsch operand, edge enhancing and edge equalization.In order to improve the accuracy of subsequent space hold information, side
The profile that edge detection algorithm obtains is intuitively important, is based on this, present invention also provides to existing a variety of edge detection algorithms
Comparative example, it is proved by many experiments that, canny operand has preferable effect, in the specific of technical scheme
During realization, canny operand pair package image to be sorted can be used and carry out edge detection, obtain objective contour image.Tool
Body may include following the description:
For Kirsch edge detection algorithm:
Kirsch is a kind of operand for detecting step shape edge, and Kirsch edge detecting operation number is for example following shown.
Above 8 templates are used to calculate the maximum response of each edge direction as edge feature value edge strength.It
Implementation method be that convolution is carried out to image respectively by above-mentioned 8 different Kirsch operands, then to obtain 8 kinds knots
Fruit carries out gradient processing again.
Laplacian edge detection algorithm:
Laplacian operand is a second-order differential operand in n dimension Euclidean space, is defined as gradient grad
Divergence div.Operation template can be used to carry out this theorem law of operation.
If f is the real function that second order can be micro-, Laplce's operand of f is defined as:
Laplce's operand of f is also all non-mixed second-order partial differential coefficient summations in Cartesian coordinates:
As a second-order differential operand, Laplce's operand is C function image to C function, for k >=2.Definition
One operand Δ: C (R) → C (R), or more generally, operand Δ: C (Ω) → C (Ω) is defined, for any
Opener Ω.
For step-like edge, there is zero crossing in marginal point in derivative, i.e. marginal point both sides second dervative takes contrary sign.According to
This takes it about the sum of x-axis direction and the second differnce in y-axis direction, is expressed as to each pixel of digital picture { f (i, j) }
Shown in following formula:
The effect for using Laplacian operand pair package (by taking mouse as an example) to be sorted to carry out edge detection goes out such as Fig. 4
It is shown.
Canny operand:
It the use of the purpose that Canny does edge detection is to find an optimal edge detection algorithm, optimal edge detection
Meaning includes:
The algorithm of good edge detection almost can be with actual edge whole in tag image.
The positioning having had, the edge of its institute's mark can be with the location of actual edge utmostly to approach, very
To consistent.
It has a smallest response, and all edges in image, all only label is primary, in this way will not fuzzy edge so that mark
Remember that result is clear, and the noise in image should not be identified as edge.
The calculus of variations is used in order to meet these requirements Canny, this method finds the method for meeting the function of specific function.
Optimal detection using exponential function item and indicate, but it closely approximates the first derivative of Gaussian function.
Canny algorithm steps:
(1) denoising
Any edge detection algorithm is impossible to handle well on untreated firsthand information, so the first step is
Convolution is made to initial data and Gaussian smoothing template, obtained image with original image compared with some slight obscure.In this way, single
An only pixel noise becomes to have little effect on the image by Gaussian smoothing.
Edge in image may be directed toward different directions, so Canny algorithm uses 4 mask detection levels, hangs down
Straight and diagonal edge.Convolution made by original image and each mask stores.Each point is marked
Know the direction at the edge of the maximum value and generation on this aspect.It thus generates in image from original image and each lights
Spend the direction of gradient map and brightness step.Following limb in the picture.Higher brightness step is relatively likely to be edge, but
It is none exact value to limit great brightness step be that edge is much, so Canny has used hysteresis threshold.Lag
Threshold value needs two threshold value-high thresholds and Low threshold.It, thus can be with assuming that the important edges in image are all continuous curves
Track the part obscured in given curve, and avoid will not no constituent curve noise pixel as edge.Can from one compared with
Big threshold value starts, this will identify the true edge for comparing and firmly believing, derived directional information before use is real from these
Edge starts to track entire edge in the picture.When tracking, using a lesser threshold value, can thus it track
The blurred portions of curve return to starting point until us.Canny algorithm is suitable for different occasions.Its parameter allows according to difference
The particular requirement of realization is adjusted to identify different local edges.Canny effect is as shown in Figure 5.
From the foregoing, it will be observed that and combine Fig. 4 and Fig. 5, Canny algorithm can be obtained package clearly to be sorted outer profile, favorably
In the subsequent area for calculating package, and then promote the accuracy of logistics package classification.
In order to reduce Edge-Detection Algorithm data volume to be processed, and the algorithm of promotion Edge-Detection Algorithm
Processing accuracy may also include and locate in advance to a series of image of package image to be sorted progress before carrying out Image Edge-Detection
Reason, specifically, package to be sorted can be carried out image gray processing, image threshold, in image filtering any one or it is any
Combination.Optionally, in a kind of specific embodiment, image gray processing, image threshold successively can be carried out to package to be sorted
Change, the processing of image filtering, is then carrying out Image Edge-Detection.
Image gray processing processing is that colored image is converted to a kind of method of gray level image, in daily engineer application
In, it is often necessary to color image is converted into gray image.Typically for collected picture, in order to accelerate processing
Cromogram is often converted to black white image by speed.Under normal circumstances, the pixel of color image is with three bytes come table
Show, each byte corresponds to the brightness value of R, G, B component, one byte representation of each pixel of the black white image after conversion
The brightness value of the point is worth between 0-255, and numerical value is bigger, and the point is whiter, i.e., brighter.Opposite value is smaller, the brightness of this point
It is lower, i.e., it is darker.Image gray processing algorithm includes maximum value process, mean value method and three kinds of weighted average method.Wherein, most
The principle of big value method is to make maximum one-component value in equal to three color components of tri- components of RGB, R=G=B=Max (R,
G, B), the image handled using maximum value process, overall brightness can improve a bit after its gray processing can be made to handle.Mean value method be than
The processing method of more common image gray processing, principle are the average value for making tri- component values of R, G, B be equal to three components,
That is R=G=B=1/3 (R+G+B), using mean value method can be it is processed after gray level image than milder.Weighted average method
Different weights is assigned respectively to R, G, B according to importance and other indexs, and makes R, G, B are equal to their weighted average,
That is R=B=G=WRR+WGG+WBB, wherein WR, WG, WB are the weight that RGB represents component respectively, when weight gets difference
Value when, weighted mean method can get the grayscale image of different gray scales.Since human eye is more sensitive to green, red is taken second place, but
It is that human eye is minimum to the susceptibility of blue, therefore, as weight WG > WR > WB, grayscale image generated is more in line with the view of human eye
Feel impression.
So-called threshold value, as separation, critical point, image threshold, also as image binaryzation, passes through one threshold of setting
Value is black by all pixels of the threshold value are less than, and all pixels greater than this threshold value regard as white.
It is the image of two kinds of colors of black and white by the image that image threshold algorithm process is crossed.Image is subjected to thresholding
The method of processing has very much, common are solid value method and double two kinds of value method admittedly.Fixed threshold method be exactly be gray level image specified one
A threshold value T, the pixel for gray value being less than given threshold value T are set to 0, and the pixel greater than threshold value T is set to 255, thus gray scale
Image is embodied as binaryzation transformation.The transforming function transformation function expression formula of fixed threshold is to be shown below:
Image by binaryzation generally contains noise, such as referring to Fig. 6, in order to improve subsequent image processing essence
Degree, can carry out denoising.
Mean filter is a kind of linear filtering algorithm, by asking target pixel points and other eight pixels around it
Average value replaces original target pixel value with this.
In this application, the effect picture that will be got by binaryzation is removed noise processed by mean filter, due to this
In place of method cannot protect the details of acquired image well, while removing noise for picture, also can largely it break
The quality of bad image can become blurred the valuable pixel being required in image unclear.The effect of mean filter is simultaneously
It is not fully up to expectations.In addition to this, the actual algorithm of mean filter, being will be other eight around each pixel in image and its
The value of a pixel, which is done, averagely to be got, although its speed in the processing for do smoothed image is fast and method is simple, but not
Useless information can be removed, the intensity of weakening noise that can only be faint can clearly find out that Gauss is filtered from the image of processing
The effect of wave, shown in effect picture such as Fig. 7 (carrying out denoising for Fig. 6):
Gaussian filtering is to be weighted and averaged processing to whole pixels contained in acquired image, obtained by final
Effect image inside each pixel value, be all pixel in its neighbouring field and itself be weighted average
Operation is got.By being scanned with a convolution or mask to acquired image, then obtains in image and own
Pixel, the weighted average of the pixel in the accurate field is obtained by above-mentioned used convolution or mask, so
The value of the pixel of replacement template center is gone with this numerical value afterwards.
The algorithm principle of gaussian filtering, actually traffic filter, its essential be acquired signal is done it is flat
Sliding processing.Digital picture is during the treatment, highly difficult in later period application, the reason is that the noise inside image is very big,
Therefore the quality of this greatly influence power image.Such issues that in order to solve, Gaussian filter, which is used to obtain signal-to-noise ratio, to be compared
High image, in this way, actual signal can be reacted.Relevant with gaussian filtering also to have Gauss-Laplace transformation, purpose is just
It is that the second best in quality edge, step are in order to obtain, first handles picture signal with Gaussian smoothing filter, noise remove is done to it
Then second dervative is sought it in processing, second order is recycled to lead zero crossing to obtain desired edge image.
Filter is exactly the mathematical model established in fact, will think that image to be processed carries out energy using this model
Conversion, noise is exactly the higher part of energy in the picture, using Gaussian smoothing, by the image data of this part high-energy
Become low energy energy part, can thus reduce the influence of noise.Gaussian filtering effect such as Fig. 8 (is carried out at denoising for Fig. 6
Reason) shown in.
Median filtering is a kind of method of nonlinear filtering in image preprocessing.Since it is to pixel a certain in image
Pixel in field sorts by its gray value, thus determine the gray value of the pixel, so, sampling window is intended to cover all
Odd number of pixels.Therefore, it can be eliminated in certain degree because of the shadow that linear filtering is obscured to image bring details
It rings, good effect is generated in terms of filtering out interference noise and scan image.During smooth noise filtering, median filtering
Play the role of extremely important obtain.Median filtering can either go filtering to remove the degree protection needs that interference noise simultaneously again can be very big
Edge acuity.The energy of holding margin signal effect of median filtering can be greatly improved by weighting flat median filtering
Power.Shown in median filtering effect such as Fig. 9 (carrying out denoising for Fig. 6).
According to each figure it is found that in a kind of specific embodiment, median filtering is can be used to by binaryzation in the application
Image carry out denoising.It certainly, can also be directly to original graph when package image to be sorted does not pass through binary conversion treatment
As either gray level image carries out denoising.
Optionally, in the image of acquisition, using suitable threshold value is set, whole image value is only left two kinds of pictures of black and white
Element, but when acquisition image, be very likely to will appear reflective phenomenon, it in this way can be to the pixel just calculated inside edge
The accuracy of point influences very big.Canny is when doing edge detection process to image, if the effect picture obtained is black back
Scape, white Foreground, then in this application, the effect image of Canny edge detection can be utilized in advance, by each column in image
Pixel ordinate maximum value and minimum value between be stuffed entirely with as white pixel point, in this way, being detected having been calculated
Within the edge contour arrived after the sum of all white pixel points, the relational expression of pixel and area is recycled, it can be in the hope of need
The true area of inclusion enclave to be classified, but due under the background of interference, needing to the figure for having not gone through removal noise
Processing as doing median filtering.
0, Figure 10 is the process signal that classification method is wrapped up in another logistics provided in an embodiment of the present invention referring to Figure 1
Figure, specifically may include the following contents:
S1001: the main view and top view of package image to be sorted are obtained.
S1002: image gray processing is carried out to main view and top view respectively, obtains corresponding gray level image.
S1003: image threshold is carried out to each gray level image, obtains corresponding black white image.
S1004: median filtering is carried out to each black white image.
S1005: the detection of side canny edge is carried out to the overhead view image after median filtering, and carries out flared end, obtains profile diagram
Picture.
S1006: the pixel inside statistics contour images, and the sum of the pixel for calculating the contour images.
S1007: according to the relational expression of the pixel and area that pre-establish, the face of package image to be sorted is calculated
Product.
S1008: canny edge detection is carried out to the main view image after median filtering, contour images is obtained, demarcates the profile
Image differs the corresponding line segment of maximum ordinate, using the height as package to be sorted.
The sum of S1009: count the pixel on the line segment and calculate the pixel.
S1010: according to the relational expression of the pixel and line segment that pre-establish, the height of package image to be sorted is calculated
Degree.
S1011: package volume corresponding with package image to be sorted is highly calculated according to area.
S1012: judging to wrap up whether volume is greater than classification thresholds, if so, executing S1013;If it is not, then executing S1014.
S1013: by package to be sorted labeled as big package.
S1014: package to be sorted is labeled as small packet.
The logistics sortation hubs package different for dimension of object, is not only very different, than big on storage mode
The article of object is placed on shelf close from the ground, and lesser logistics package can be placed on from the farther away shelf in the inside,
And large-sized object and also different in terms of transport, such as means of transport are just different, and big object can be transported with fork truck
It send, small object can be transported with conveyer belt etc., these all and by the object classification of size dimension have a very large relationship, at this
In inventive embodiments, the classification effect in real logistics operation is emulated, can be more than certain threshold value (such as 120cm by size3)
Object carries out the label of classification, i.e. object tag small packet printed words of the size within a certain threshold value in the picture, and will be more than should
The big package printed words of the object tag of threshold size.It, may will be many of different specifications and sizes it is, of course, also possible to multiple threshold values are arranged
Object segments multiple classifications, is then marked accordingly to them again, to meet in reality, there are many kinds of size rulers
Very little widely different object package, the requirement to classification.
When specifically being marked, English alphabet can be used, alphabets consisting in Chinese can also be used, identified currently using any
The other tag mark of the tag class of package is identified, and the application does not do any restriction to this.
Based on the above embodiment, in order to make those skilled in the art that the technical solution of the application be more clearly understood, this Shen
Please by taking package to be sorted is mouse as an example, the realization process of the technical solution of the application is illustrated, specifically can include:
After carrying out S1001-S1007 to mouse images, the sum of pixel of contoured interior of top view of mouse is
35184, the front face area for obtaining mouse is 45.4438464cm2.Very little is differed with real area error.
After by S1008-S1010, the height calibration result of main view is seen shown in Figure 11, mouse side
Height is about 2.60cm, and volume is about 118.13cm3, and actually it is consistent.
When the classification thresholds set is 120cm3, mouse can be labeled as small packet.
From the foregoing, it will be observed that the embodiment of the present invention is greatly reduced by carrying out series of preprocessing to package image to be sorted
The treating capacity of later data, improves the picture quality of package to be sorted, to be conducive to improve entire logistics package point
The working efficiency of class and classification accuracy.
It, can also be upper in order to further increase the efficiency of subsequent logistics sorting and transport in another embodiment
It states based on after package dimensions classification, of a sort package can be continued to carry out secondary classification according to destination address.And it is specific
How to be classified according to destination address, sees existing any method and do not repeating herein.
The embodiment of the present invention provides corresponding realization device also directed to logistics package classification method, further such that described
Method has more practicability.Logistics provided in an embodiment of the present invention package sorter is introduced below, it is described below
Sorter is wrapped up in logistics can correspond to each other reference with above-described logistics package classification method.
Referring to Figure 12, Figure 12 is that logistics provided in an embodiment of the present invention wraps up sorter under a kind of specific embodiment
Structure chart, the device can include:
Image capture module 1201, for obtaining the package image to be sorted under default shooting angle and shooting background;
Edge detection module 1202 obtains package image to be sorted for carrying out edge detection to package image to be sorted
Objective contour image;
Pixel computing module 1203, for calculating the sum of pixel in objective contour image;
Package dimensions computing module 1204 is calculated for the relational expression according to the pixel and geometric element pre-established
The space hold value of package image to be sorted is obtained, using as the standard classified to package to be sorted.
Optionally, in some embodiments of the present embodiment, the package dimensions computing module 1204 can also include:
Areal calculation submodule is top view for package image to be sorted, calculates the inside of the corresponding profile of top view
The area of package image to be sorted is calculated according to the relational expression of the pixel and area that pre-establish in the sum of pixel;
Height computational submodule is main view for package image to be sorted, calculates the sum of pixel of target segment, root
According to the relational expression of the pixel and length that pre-establish, the height of package image to be sorted is calculated;Target segment is main view
The corresponding line segment of the maximum height on y direction of the contour images of figure;
Volume computational submodule, the body for package corresponding with package image to be sorted is highly calculated according to area
Product.
Optionally, in other embodiments of the present embodiment, Figure 13 is please referred to, it is pre- that described device may also include image
Processing module 1205, for any one in package image progress image gray processing to be sorted, image threshold, image filtering
Kind or any combination.
Specifically, may also include package categorization module 1206, the package categorization module 1206 is for example can include:
Threshold value acquisition submodule, for obtaining package classification thresholds;
Judging submodule, for judging whether space hold value is greater than classification thresholds;
Implementation sub-module is judged, for judging that space hold value is greater than classification thresholds, then by package to be sorted labeled as big
Package;Judge that space hold value is less than or equal to classification thresholds, then package to be sorted is labeled as small packet.
The function of each functional module of the package sorter of logistics described in the embodiment of the present invention can be implemented according to the above method
Method specific implementation in example, specific implementation process are referred to the associated description of above method embodiment, no longer superfluous herein
It states.
From the foregoing, it will be observed that the embodiment of the present invention improves the accuracy and classification effectiveness of logistics package classification.
The embodiment of the invention also provides a kind of logistics to wrap up sorting device, specifically can include:
Memory, for storing computer program;
Processor realizes logistics package classification method described in any one embodiment as above for executing computer program
Step.
The function of each functional module of the package sorting device of logistics described in the embodiment of the present invention can be implemented according to the above method
Method specific implementation in example, specific implementation process are referred to the associated description of above method embodiment, no longer superfluous herein
It states.
From the foregoing, it will be observed that the embodiment of the present invention improves the accuracy and classification effectiveness of logistics package classification.
The embodiment of the invention also provides a kind of computer readable storage mediums, are stored with logistics package sort program, institute
State logistics package sort program when being executed by processor as above logistics package classification method described in any one embodiment the step of.
The function of each functional module of computer readable storage medium described in the embodiment of the present invention can be according to above method reality
The method specific implementation in example is applied, specific implementation process is referred to the associated description of above method embodiment, herein no longer
It repeats.
From the foregoing, it will be observed that the embodiment of the present invention improves the accuracy and classification effectiveness of logistics package classification.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with it is other
The difference of embodiment, same or similar part may refer to each other between each embodiment.For being filled disclosed in embodiment
For setting, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is referring to method part
Explanation.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Above to a kind of logistics package classification method, device, equipment and computer-readable storage medium provided by the present invention
Matter is described in detail.Used herein a specific example illustrates the principle and implementation of the invention, above
The explanation of embodiment is merely used to help understand method and its core concept of the invention.It should be pointed out that for the art
Those of ordinary skill for, without departing from the principle of the present invention, can also to the present invention carry out it is several improvement and repair
Decorations, these improvements and modifications also fall within the scope of protection of the claims of the present invention.
Claims (10)
1. classification method is wrapped up in a kind of logistics characterized by comprising
Obtain the package image to be sorted under default shooting angle and shooting background;
Edge detection is carried out to the package image to be sorted, obtains the objective contour image of the package image to be sorted;
The sum of pixel in the objective contour image is calculated, according to the relational expression of the pixel and geometric element that pre-establish,
The space hold value of the package image to be sorted is calculated, using as the standard classified to the package to be sorted.
2. classification method is wrapped up in logistics according to claim 1, which is characterized in that the objective contour image includes main view
The target segment of figure and the profile of top view, it is described to calculate the sum of pixel in the objective contour image, according to pre-establishing
Pixel and geometric element relational expression, be calculated it is described it is to be sorted package image space hold value include:
The package image to be sorted is top view, calculates the sum of interior pixels point of the corresponding profile of the top view, according to
The area of the package image to be sorted is calculated in the relational expression of the pixel and area that pre-establish;
The package image to be sorted is main view, the sum of the pixel of the target segment is calculated, according to the picture pre-established
The height of the package image to be sorted is calculated in the relational expression of vegetarian refreshments and length;The target segment is the main view
Contour images on y direction the corresponding line segment of maximum height;
The volume of the corresponding package of the package image to be sorted is calculated according to the area and the height.
3. classification method is wrapped up in logistics according to claim 1, which is characterized in that it is described obtain default shooting angle and
After package image to be sorted under shooting background, further includes:
Image preprocessing is carried out to the package image to be sorted;
It is described to include: to the package image progress image preprocessing to be sorted
Image gray processing, image threshold, any one or any combination in image filtering are carried out to the package to be sorted.
4. wrapping up classification method to logistics described in 3 any one according to claim 1, which is characterized in that the basis is preparatory
The pixel of foundation and the relational expression of geometric element include:
Obtain the figure of the known dimensions for the predetermined number drawn in default background;
Record the default line segment length of each figure and area and the corresponding pixel of corresponding number of pixels and each figure
Number, as two groups of test datas;
Two groups of test datas are fitted respectively using least square method, obtain the relational expression and picture of pixel and length
The relational expression of vegetarian refreshments and area.
5. classification method is wrapped up in logistics according to claim 4, which is characterized in that in the pixel that the basis pre-establishes
The relational expression of point and geometric element is calculated after the space hold value of the package image to be sorted, further includes:
Obtain package classification thresholds;
Judge whether the space hold value is greater than the classification thresholds;
If so, by the package to be sorted labeled as big package;
If it is not, the package to be sorted is then labeled as small packet.
6. sorter is wrapped up in a kind of logistics characterized by comprising
Image capture module, for obtaining the package image to be sorted under default shooting angle and shooting background;
Edge detection module obtains the package image to be sorted for carrying out edge detection to the package image to be sorted
Objective contour image;
Pixel computing module, for calculating the sum of pixel in the objective contour image;
Package dimensions computing module is calculated described for the relational expression according to the pixel and geometric element pre-established
The space hold value of package image to be sorted, using as the standard classified to the package to be sorted.
7. sorter is wrapped up in logistics according to claim 6, which is characterized in that the package dimensions computing module packet
It includes:
Areal calculation submodule is top view for the package image to be sorted, calculates the corresponding profile of the top view
The package image to be sorted is calculated according to the relational expression of the pixel and area that pre-establish in the sum of interior pixels point
Area;
Height computational submodule is main view for the package image to be sorted, calculates the sum of pixel of target segment, root
According to the relational expression of the pixel and length that pre-establish, the height of the package image to be sorted is calculated;The score
Section is the corresponding line segment of the maximum height on y direction of the contour images of the main view;
Volume computational submodule, it is corresponding for the package image to be sorted to be calculated according to the area and the height
The volume of package.
8. sorter is wrapped up in logistics according to claim 7, which is characterized in that further include image pre-processing module, use
In carrying out image gray processing, image threshold, any one or any group in image filtering to the package image to be sorted
It closes.
9. sorting device is wrapped up in a kind of logistics, which is characterized in that including processor, the processor is deposited for executing in memory
The step of classification method is wrapped up in logistics as described in any one of claim 1 to 5 is realized when the computer program of storage.
10. a kind of computer readable storage medium, which is characterized in that be stored with logistics packet on the computer readable storage medium
Sort program is wrapped up in, the logistics package sort program realizes the logistics as described in any one of claim 1 to 5 when being executed by processor
The step of wrapping up classification method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810989007.9A CN109190696A (en) | 2018-08-28 | 2018-08-28 | A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810989007.9A CN109190696A (en) | 2018-08-28 | 2018-08-28 | A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109190696A true CN109190696A (en) | 2019-01-11 |
Family
ID=64916440
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810989007.9A Pending CN109190696A (en) | 2018-08-28 | 2018-08-28 | A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109190696A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110502966A (en) * | 2019-07-01 | 2019-11-26 | 广州市川流信息科技有限公司 | The classification information of package obtains equipment, method and storage device |
CN110721916A (en) * | 2019-10-18 | 2020-01-24 | 兴盛社区网络服务股份有限公司 | Logistics sorting system and sorting method thereof |
CN111860136A (en) * | 2020-06-08 | 2020-10-30 | 北京阿丘机器人科技有限公司 | Parcel positioning method, device, equipment and computer readable storage medium |
CN112232335A (en) * | 2019-07-15 | 2021-01-15 | 德国邮政股份公司 | Determination of distribution and/or sorting information for the automated distribution and/or sorting of mailpieces |
CN112991423A (en) * | 2021-03-15 | 2021-06-18 | 上海东普信息科技有限公司 | Logistics package classification method, device, equipment and storage medium |
CN113252103A (en) * | 2021-05-11 | 2021-08-13 | 安徽理工大学 | Method for calculating volume and mass of material pile based on MATLAB image recognition technology |
CN114758250A (en) * | 2022-06-15 | 2022-07-15 | 山东青岛烟草有限公司 | Full-specification flexible automatic sorting control method and device based on artificial intelligence |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5137362A (en) * | 1990-03-26 | 1992-08-11 | Motorola, Inc. | Automatic package inspection method |
CN103630074A (en) * | 2013-11-29 | 2014-03-12 | 北京京东尚科信息技术有限公司 | Method and device for measuring minimum package volume of object |
CN104501718A (en) * | 2014-12-18 | 2015-04-08 | 江苏物联网研究发展中心 | Parcel size measuring device based on visual sense |
CN105466534A (en) * | 2015-11-13 | 2016-04-06 | 广州市中崎商业机器股份有限公司 | Logistic freight automatic calculating device and calculating method thereof |
CN206168812U (en) * | 2016-09-27 | 2017-05-17 | 大连海洋大学 | Thick sorter of express delivery |
CN206578035U (en) * | 2017-02-07 | 2017-10-24 | 申佳君 | A kind of express delivery categorizing system based on internet |
CN107388960A (en) * | 2016-05-16 | 2017-11-24 | 杭州海康机器人技术有限公司 | A kind of method and device for determining object volume |
CN108240793A (en) * | 2018-01-26 | 2018-07-03 | 广东美的智能机器人有限公司 | Dimension of object measuring method, device and system |
CN108416804A (en) * | 2018-02-11 | 2018-08-17 | 深圳市优博讯科技股份有限公司 | Obtain method, apparatus, terminal device and the storage medium of target object volume |
-
2018
- 2018-08-28 CN CN201810989007.9A patent/CN109190696A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5137362A (en) * | 1990-03-26 | 1992-08-11 | Motorola, Inc. | Automatic package inspection method |
CN103630074A (en) * | 2013-11-29 | 2014-03-12 | 北京京东尚科信息技术有限公司 | Method and device for measuring minimum package volume of object |
CN104501718A (en) * | 2014-12-18 | 2015-04-08 | 江苏物联网研究发展中心 | Parcel size measuring device based on visual sense |
CN105466534A (en) * | 2015-11-13 | 2016-04-06 | 广州市中崎商业机器股份有限公司 | Logistic freight automatic calculating device and calculating method thereof |
CN107388960A (en) * | 2016-05-16 | 2017-11-24 | 杭州海康机器人技术有限公司 | A kind of method and device for determining object volume |
CN206168812U (en) * | 2016-09-27 | 2017-05-17 | 大连海洋大学 | Thick sorter of express delivery |
CN206578035U (en) * | 2017-02-07 | 2017-10-24 | 申佳君 | A kind of express delivery categorizing system based on internet |
CN108240793A (en) * | 2018-01-26 | 2018-07-03 | 广东美的智能机器人有限公司 | Dimension of object measuring method, device and system |
CN108416804A (en) * | 2018-02-11 | 2018-08-17 | 深圳市优博讯科技股份有限公司 | Obtain method, apparatus, terminal device and the storage medium of target object volume |
Non-Patent Citations (1)
Title |
---|
宓逸舟: "基于双目视觉的快递包裹体积计量系统", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110502966A (en) * | 2019-07-01 | 2019-11-26 | 广州市川流信息科技有限公司 | The classification information of package obtains equipment, method and storage device |
CN110502966B (en) * | 2019-07-01 | 2023-06-30 | 广州市川流信息科技有限公司 | Classified information acquisition equipment, method and storage device for packages |
CN112232335A (en) * | 2019-07-15 | 2021-01-15 | 德国邮政股份公司 | Determination of distribution and/or sorting information for the automated distribution and/or sorting of mailpieces |
CN110721916A (en) * | 2019-10-18 | 2020-01-24 | 兴盛社区网络服务股份有限公司 | Logistics sorting system and sorting method thereof |
CN110721916B (en) * | 2019-10-18 | 2021-05-07 | 兴盛社区网络服务股份有限公司 | Logistics sorting system and sorting method thereof |
CN111860136A (en) * | 2020-06-08 | 2020-10-30 | 北京阿丘机器人科技有限公司 | Parcel positioning method, device, equipment and computer readable storage medium |
CN111860136B (en) * | 2020-06-08 | 2024-03-29 | 北京阿丘机器人科技有限公司 | Package positioning method, device, equipment and computer readable storage medium |
CN112991423A (en) * | 2021-03-15 | 2021-06-18 | 上海东普信息科技有限公司 | Logistics package classification method, device, equipment and storage medium |
CN113252103A (en) * | 2021-05-11 | 2021-08-13 | 安徽理工大学 | Method for calculating volume and mass of material pile based on MATLAB image recognition technology |
CN114758250A (en) * | 2022-06-15 | 2022-07-15 | 山东青岛烟草有限公司 | Full-specification flexible automatic sorting control method and device based on artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109190696A (en) | A kind of logistics package classification method, device, equipment and readable storage medium storing program for executing | |
CN109342456B (en) | Welding spot defect detection method, device and equipment and readable storage medium | |
CN108230321A (en) | Defect inspection method and device | |
CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
CN108636830A (en) | The method, apparatus and equipment of defective capsule detection sorting based on machine vision | |
CN106951869B (en) | A kind of living body verification method and equipment | |
CN107766861A (en) | The recognition methods of character image clothing color, device and electronic equipment | |
CN110427932A (en) | The method and device of multiple document fields in a kind of identification image | |
CN110751620B (en) | Method for estimating volume and weight, electronic device, and computer-readable storage medium | |
CN108647634A (en) | Framing mask lookup method, device, computer equipment and storage medium | |
CN108428214A (en) | A kind of image processing method and device | |
CN108765584A (en) | Laser point cloud data collection augmentation method, apparatus and readable storage medium storing program for executing | |
CN110570442A (en) | Contour detection method under complex background, terminal device and storage medium | |
CN104574312A (en) | Method and device of calculating center of circle for target image | |
CN103679169A (en) | Method and device for extracting image features | |
CN111860369A (en) | Fraud identification method and device and storage medium | |
CN109685142A (en) | A kind of image matching method and device | |
CN109948521A (en) | Image correcting error method and device, equipment and storage medium | |
CN112132812A (en) | Certificate checking method and device, electronic equipment and medium | |
CN112417931B (en) | Method for detecting and classifying water surface objects based on visual saliency | |
CN107507130A (en) | A kind of quickly QFN chip pins image obtains and amplification method | |
CN107845120A (en) | PET image reconstruction method, system, terminal and readable storage medium storing program for executing | |
CN110440792A (en) | Navigation information extracting method based on small feature loss degree of irregularity | |
CN116342519A (en) | Image processing method based on machine learning | |
CN105354823A (en) | Tree-ring image edge extraction and segmentation system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190111 |
|
RJ01 | Rejection of invention patent application after publication |