CN109285154A - A kind of method and device detecting the stone age - Google Patents
A kind of method and device detecting the stone age Download PDFInfo
- Publication number
- CN109285154A CN109285154A CN201811161771.3A CN201811161771A CN109285154A CN 109285154 A CN109285154 A CN 109285154A CN 201811161771 A CN201811161771 A CN 201811161771A CN 109285154 A CN109285154 A CN 109285154A
- Authority
- CN
- China
- Prior art keywords
- stone age
- feature extraction
- coordinate
- pisiform
- piece
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30008—Bone
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the present application provides a kind of method and device for detecting the stone age, it is related to machine learning techniques field, this method comprises: obtaining the stone age piece of user to be detected, and each bone in stone age piece is adjusted to base position, carpal area is determined from stone age piece, then using the pisiform in identification model identification carpal area, identification model is using the stone age piece for marking the carpal area of pisiform in advance as training sample, it is determined after being trained to depth residual error network, the stone age of user to be detected is finally determined according to the recognition result of pisiform.Due in carpal area, pisiform goes out specific stage stone age of representing, therefore in the embodiment of the present application, the pisiform in carpal area is individually identified using identification model, the stone age of user to be detected is determined according to the recognition result of pisiform, it does not need human subjective and the stone age is judged according to stone age piece, on the one hand improve the precision of detection stone age, on the other hand improve the efficiency of detection stone age.
Description
Technical field
The present embodiments relate to machine learning techniques field more particularly to a kind of method and devices for detecting the stone age.
Background technique
" stone age " is the abbreviation of bone age, is that youngsters and children skeleton development level is obtained with bone development standard comparing
Developmental age, it more can accurately reflect the maturity of body than age, height, weight, more accurately reflection
The growth level and maturity of body.
The biological age for carrying out interpretation children clinically by the detection stone age, passes through the difference of biological age and calendar age
It assesses child development situation, understands the sexually matured trend of children, predict the adult height etc. of children, and be widely used in and influence youngster
The Treatment monitoring of virgin growth and development disease has very great help to the diagnosis of some paediatrics endocrine system diseases.
It is main by shooting stone age piece in the prior art, by manually checking that stone age piece estimates the stone age.This method relies on artificial
The stone age is estimated according to stone age piece, and precision is big by the subjective factor image of people, and efficiency is lower.
Summary of the invention
The stone age is manually estimated according to stone age piece due to relying in the prior art, precision is big by the subjective factor image of people, effect
The lower problem of rate, the embodiment of the present application provide a kind of method and device for detecting the stone age.
In a first aspect, the embodiment of the present application provides a kind of method for detecting the stone age, comprising:
The stone age piece of user to be detected is obtained, and adjusts each bone in the stone age piece to base position；
Carpal area is determined from the stone age piece；
Pisiform in the carpal area is identified using identification model, the identification model is to mark pisiform in advance
Carpal area stone age piece be training sample, after being trained to depth residual error network determine；
The stone age of the user to be detected is determined according to the recognition result of the pisiform.
It is optionally, described that carpal area is determined from the stone age piece, comprising:
The coordinate of the corresponding key point of carpal area in the stone age piece, the parted pattern are determined using parted pattern is
Using mark in advance key point coordinate multiple stone age pieces as training sample, after being trained to depth residual error network determine；
The carpal area is determined according to the coordinate of the corresponding key point of the carpal area.
Optionally, the pisiform identified using identification model in the carpal area, comprising:
The stone age piece of the carpal area is handled through N number of continuous convolution feature extraction block, obtains the carpal area
Characteristics of image, N is greater than 0, and the convolution feature extraction block includes L convolution module, and L is greater than 0, in any one convolution module
Including convolutional layer, BN layers and excitation layer；For the continuous first convolution feature of any two in N number of convolution feature extraction block
Extract block and the second convolution feature extraction block, the second characteristics of image and described first of the second convolution feature extraction block output
Convolution feature extraction block output the first characteristics of image be added after as third product feature extraction block input or it is N number of continuously
The output of convolution feature extraction block；The third convolution feature extraction block be after the second convolution feature extraction block and
With the continuous convolution feature extraction block of the second convolution feature extraction block；
The characteristics of image of the carpal area is inputted into full articulamentum, exports the recognition result of the pisiform.
Optionally, each bone in the adjustment stone age piece is to base position, comprising:
Obtain the coordinate of preset reference point；
The coordinate of key point in the stone age piece is determined using adjustment model, the adjustment model is to mark key in advance
Multiple stone age pieces of the coordinate of point are training sample, are determined after being trained to depth residual error network, the seat of the key point
The coordinate of mark and the preset reference point is the coordinate in the same coordinate system；
The current of each bone of stone age piece is determined according to the coordinate of the coordinate of the preset reference point and the key point
Corresponding relationship between position and base position；
Each bone in the stone age piece is adjusted to base position according to the corresponding relationship.
Optionally, the recognition result according to the pisiform determines the stone age of the user to be detected, comprising:
When including pisiform in determining the carpal area, determine that the stone age of the user to be detected is first default
Value；
When not including pisiform in determining the carpal area, determine that the stone age of the user to be detected is second default
Value.
Second aspect, the embodiment of the present application provide a kind of device for detecting the stone age, comprising:
Module is obtained, for obtaining the stone age piece of user to be detected, and adjusts each bone in the stone age piece to benchmark
Position；
Divide module, for determining carpal area from the stone age piece；
Identification module, for identifying the pisiform in the carpal area using identification model, the identification model be with
The stone age piece of the carpal area of label pisiform is training sample in advance, is determined after being trained to depth residual error network；
Detection module, for determining the stone age of the user to be detected according to the recognition result of the pisiform.
Optionally, the segmentation module is specifically used for:
The coordinate of the corresponding key point of carpal area in the stone age piece, the parted pattern are determined using parted pattern is
Using mark in advance key point coordinate multiple stone age pieces as training sample, after being trained to depth residual error network determine；
The carpal area is determined according to the coordinate of the corresponding key point of the carpal area.
Optionally, the identification module is specifically used for:
The stone age piece of the carpal area is handled through N number of continuous convolution feature extraction block, obtains the carpal area
Characteristics of image, N is greater than 0, and the convolution feature extraction block includes L convolution module, and L is greater than 0, in any one convolution module
Including convolutional layer, BN layers and excitation layer；For the continuous first convolution feature of any two in N number of convolution feature extraction block
Extract block and the second convolution feature extraction block, the second characteristics of image and described first of the second convolution feature extraction block output
Convolution feature extraction block output the first characteristics of image be added after as third product feature extraction block input or it is N number of continuously
The output of convolution feature extraction block；The third convolution feature extraction block be after the second convolution feature extraction block and
With the continuous convolution feature extraction block of the second convolution feature extraction block；
The characteristics of image of the carpal area is inputted into full articulamentum, exports the recognition result of the pisiform.
Optionally, the acquisition module is specifically used for:
Obtain the coordinate of preset reference point；
The coordinate of key point in the stone age piece is determined using adjustment model, the adjustment model is to mark key in advance
Multiple stone age pieces of the coordinate of point are training sample, are determined after being trained to depth residual error network, the seat of the key point
The coordinate of mark and the preset reference point is the coordinate in the same coordinate system；
The current of each bone of stone age piece is determined according to the coordinate of the coordinate of the preset reference point and the key point
Corresponding relationship between position and base position；
Each bone in the stone age piece is adjusted to base position according to the corresponding relationship.
Optionally, the detection module is specifically used for:
When including pisiform in determining the carpal area, determine that the stone age of the user to be detected is first default
Value；
When not including pisiform in determining the carpal area, determine that the stone age of the user to be detected is second default
Value.
The third aspect, the embodiment of the present application provide it is a kind of detect the stone age equipment, including at least one processor and
At least one processor, wherein the storage unit is stored with computer program, when described program is executed by the processor
When, so that the step of processor executes first aspect the method.
Fourth aspect, the embodiment of the present application provide a kind of computer-readable medium, and being stored with can be by the detection stone age
The computer program that equipment executes, when described program is run in the equipment for detecting the stone age, so that the detection stone age sets
It is standby to execute the step of first aspect states method.
In the embodiment of the present application, since in carpal area, pisiform goes out to represent specific stage stone age, because
In this embodiment of the present application, the pisiform in carpal area is individually identified using identification model, according to the recognition result of pisiform
The stone age for determining user to be detected does not need human subjective according to stone age piece and judges the stone age, on the one hand improve detection bone
On the other hand the precision in age improves the efficiency of detection stone age.Secondly, pisiform can be Chong Die with ostriquetrum after growing, therefore examining
When the quantity of ossification centre determines the stone age in survey carpal area, pisiform and ostriquetrum may be identified as one piece of bone, because
This, identification pisiform can determine that the stone age supplements to the quantity by ossification centre in detection of carpal bone region, improve detection bone
The precision in age.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill in field, without any creative labor, it can also be obtained according to these attached drawings
His attached drawing.
Fig. 1 is a kind of flow diagram of method for detecting the stone age provided by the embodiments of the present application；
Fig. 2 is a kind of schematic diagram of stone age piece provided by the embodiments of the present application；
Fig. 3 is a kind of schematic diagram of carpal area provided by the embodiments of the present application；
Fig. 4 is a kind of flow diagram of method for adjusting stone age piece provided by the embodiments of the present application；
Fig. 5 is a kind of structural schematic diagram of depth residual error network provided by the embodiments of the present application；
Fig. 6 is a kind of structural schematic diagram of convolution feature extraction block provided by the embodiments of the present application；
Fig. 7 is a kind of flow diagram of method for identifying pisiform provided by the embodiments of the present application；
Fig. 8 is a kind of structural schematic diagram of device for detecting the stone age provided by the embodiments of the present application；
Fig. 9 is a kind of structural schematic diagram of equipment for detecting the stone age provided by the embodiments of the present application.
Specific embodiment
In order to which the purpose of the present invention, technical solution and beneficial effect is more clearly understood, below in conjunction with attached drawing and implementation
Example, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only used to explain this hair
It is bright, it is not intended to limit the present invention.
Stone age: the abbreviation of bone age needs the specific image by means of bone in X-ray camera shooting to determine.Usually want
The X-ray of people's left hand wrist portion is shot, doctor is observed in left hand metacarpal-phalangeal, carpal bone and the ossification of radioulna lower end by X-ray
The development degree of the heart, to determine the stone age.
The technical solution that the stone age is detected in the embodiment of the present application is suitable for hospital and detects the teen-age stone age, determines for doctor
Adolescent growth developmental state provides reference, while helping diagnosis chromosome abnormality, genetic disease, endocrine system disease etc.
Disease.
Fig. 1 illustrates a kind of flow diagram of method for detecting the stone age provided by the embodiments of the present application, the stream
Journey can be executed by the device for detecting the stone age, specifically includes the following steps:
Step S101, obtains the stone age piece of user to be detected, and adjusts each bone in stone age piece to base position.
Stone age piece refers to that the specific image shot using X-ray, the usually X-ray of shooting people's left hand are exemplary as stone age piece
Ground, as shown in Figure 2.
After the stone age piece for obtaining user to be detected, stone age piece is pre-processed, preprocessing process mainly includes body of gland
Segmentation and image normalization.
Body of gland segmentation is the following steps are included: use gaussian filtering to whole stone age piece first, by the result binaryzation of filtering,
The threshold value of binaryzation is by asking the maximum kind distance method of image grey level histogram to obtain.Then after the result of binaryzation being expanded
Independent region unit one by one is obtained by unrestrained water law (flood fill), counts the area of each region unit.By area maximum
Region unit retain, the hand bone image split.The hand bone image being partitioned into is pasted to the length and width of one He hand bone image
On the pure black image being consistent.
Image normalization is the following steps are included: stone age piece is the image of dicom format, first according to dicom information, selection window
Wide window position switchs to the stone age picture of png format.By on the upside of stone age picture or two sides add black surround, by stone age piece figure
As length-width ratio is adjusted to 1:1, stone age picture is finally zoomed into 512*512 size.
Step S102 determines carpal area from stone age piece.
Carpal area is located at wrist portion, is made of 8 pieces of ossiculums, is arranged in two rows, it is hand boat that nearside, which is arranged from the lateral ulnar side of oar,
Bone, lunar, ostriquetrum and pisiform both participate in the composition of articulatio radiocarpea in addition to pisiform.It is big that distal side, which is arranged from the lateral ulnar side of oar,
Multangulum, lesser trapezium bone, capitatum and hamate bone both participate in the composition of articulatio carpometacarpicus communis, illustratively, as shown in Figure 3.
Step S103, using the pisiform in identification model identification carpal area.
In carpal bone growth course, the appearance of pisiform corresponding certain stage stone age.Since pisiform is Chong Die with ostriquetrum,
Therefore in ossification centre's quantity in detection of carpal bone region, pisiform and ostriquetrum can be identified as to an ossification centre, detected
Carpal area in the quantity of ossification centre at most may be 7, and in carpal area at most may include 8 ossification centres.
By identifying pisiform, the method for detecting the stone age by ossification centre's quantity can effectively be supplemented.
Identification model is using the stone age piece for marking the carpal area of pisiform in advance as training sample, to depth residual error network
It is determined after being trained.
Step S104 determines the stone age of user to be detected according to the recognition result of pisiform.
Optionally, the recognition result of pisiform includes in carpal area comprising not including pea in pisiform and carpal area
Bone.The presence or absence of pisiform corresponds to two stages stone age, when including pisiform in determining carpal area, determines user's to be detected
Stone age is the first preset value.When not including pisiform in determining carpal area, determine that the stone age of user to be detected is second pre-
If value.Illustratively, when in determining carpal area comprising pisiform, determine that the stone age of user to be detected is 10 years old or more.?
When determining in carpal area not comprising pisiform, determine that the stone age of user to be detected is 10 years old or less.
Since in carpal area, pisiform goes out to represent specific stage stone age, therefore the embodiment of the present application
In, the pisiform in carpal area is individually identified using identification model, user to be detected is determined according to the recognition result of pisiform
Stone age, do not need human subjective according to stone age piece and judge the stone age, thus on the one hand improve detection the stone age precision, another party
Face improves the efficiency of detection stone age.Secondly, pisiform can be Chong Die with ostriquetrum after growing, therefore ossify in detection of carpal bone region
When the quantity at center determines the stone age, pisiform and ostriquetrum may be identified as one piece of bone, and therefore, identification pisiform can be right
It determines that the stone age is supplemented by the quantity of ossification centre in detection of carpal bone region, improves the precision of detection stone age.
Optionally, in above-mentioned steps S101, each bone in stone age piece is adjusted to base position, specifically includes following step
Suddenly, as shown in Figure 4:
Step S401 obtains the coordinate of preset reference point.
The coordinate of preset reference point can be the coordinate of preset part bone, for indicating the benchmark position of part bone
It sets, for example the coordinate of preset reference point can be the preset coordinate of point relevant to middle finger, for indicating the base position of middle finger,
The coordinate of preset reference point is also possible to the preset coordinate of point relevant to little finger of toe, for indicating the base position of little finger of toe.
Step S402 determines that the coordinate of key point in stone age piece, adjustment model are with preparatory label pass using adjustment model
Multiple stone age pieces of the coordinate of key point are training sample, are determined after being trained to depth residual error network.
The coordinate of coordinate and the preset reference point of key point is the coordinate in the same coordinate system.
Illustratively, multiple stone age pieces are obtained, then the key point in handmarking's stone age piece around little finger of toe, it then will mark
The stone age piece input depth residual error network of note key point is trained, when the objective function of depth residual error network meets preset condition
When, determine adjustment model.When obtaining the stone age piece of user to be detected, stone age piece is inputted into adjustment model, is determined in stone age piece
Key point around little finger of toe.
Step S403 determines the current location of each bone of stone age piece according to the coordinate of the coordinate of preset reference point and key point
Corresponding relationship between base position.
Step S404 adjusts each bone in stone age piece to base position according to corresponding relationship.
When preset reference point and key point are all relevant to little finger of toe, according to the coordinate and key point of preset reference point
Coordinate can determine corresponding pass of the little finger of toe in current location and the little finger of toe in stone age piece between the base position in stone age piece
System, the further also corresponding relationship in available stone age piece between the current location and base position of other bones, wherein right
It should be related to including translation relation and rotation relationship.Then each bone in stone age piece is adjusted to benchmark position according to corresponding relationship
It sets.Before determining carpal area, first each bone in the stone age piece of user to be detected is adjusted to base position, to improve
The precision for determining carpal area further improves the precision of detection stone age.
Optionally, in above-mentioned steps S102, carpal bone area can be determined from unadjusted stone age piece using parted pattern
Domain can also determine carpal area using parted pattern, wherein adjustment stone age piece refers to the adjustment stone age from stone age piece adjusted
Each bone in piece is to base position.Parted pattern is to mark multiple stone age pieces of the coordinate of key point for training sample in advance
This, determines after being trained to depth residual error network.
Lower mask body introduces the training process of parted pattern: obtaining multiple stone age pieces as training sample.For each
Stone age piece pre-processes stone age piece, then each bone in stone age piece is adjusted to base position, wherein adjustment stone age piece
Position and pretreated process carried out to stone age piece described above, details are not described herein again.By mark personnel every
It opens and marks key point in stone age piece, key point is the point in stone age piece near carpal area.Then data are carried out to training sample
Data volume is enhanced to 10 times of original data volume by enhancing, and the method for data enhancing includes but is not limited to:
1, Random-Rotation certain angle.
2, random to translate 0~30 pixel up and down.
3,0.85~1.15 times is scaled at random.
4, picture contrast and brightness are shaken on a small quantity.
Training sample input depth residual error network is trained again later.According to the coordinate of the key point of mark when training
Loss function is calculated with the coordinate of the key point of neural network forecast, by the method training of backpropagation, trained optimization algorithm makes
With the sgd algorithm with momentum and ladder decaying.
Further, after training parted pattern, the corresponding key of carpal area in stone age piece is determined using parted pattern
The coordinate of point, determines carpal area according to the coordinate of the corresponding key point of carpal area.
Due to first being divided from stone age piece using parted pattern when identifying the pisiform in carpal area using identification model
Carpal area is cut out, detection range is reduced, then identifies the pisiform in carpal area again, the stone age is determined according to recognition result,
To improve the precision of detection stone age.
Optionally, in above-mentioned steps S103, the training process of identification model is as follows:
The stone age piece for the carpal area being partitioned into from multiple stone age pieces is obtained as training sample, wherein a part of carpal bone
The stone age piece in region does not include pisiform, and the stone age piece of another part carpal area includes pisiform, for each carpal bone area
The stone age piece in domain marks identification types by mark personnel in the stone age piece of carpal area, identification types include pisiform and
Without two kinds of pisiform.Then data enhancing is carried out to training sample, the method for data enhancing includes but is not limited to:
1, Random-Rotation certain angle.
2, random to translate 0~30 pixel up and down.
3,0.85~1.15 times is scaled at random.
4, picture contrast and brightness are shaken on a small quantity.
Training sample input depth residual error network is trained again later.According to the identification types and net of mark when training
The identification types of network prediction calculate loss function, and by the method training of backpropagation, trained optimization algorithm is used with dynamic
The sgd algorithm of amount and ladder decaying.
Optionally, above-mentioned depth residual error network structure as shown in figure 5, include N number of continuous convolution feature extraction block with
And a full articulamentum, for the continuous first convolution feature extraction block of any two and second in N number of convolution feature extraction block
Convolution feature extraction block, what the second characteristics of image and the first convolution feature extraction block of the second convolution feature extraction block output exported
First characteristics of image be added after as third accumulate feature extraction block input or N number of continuous convolution feature extraction block it is defeated
Out.Third convolution feature extraction block is after the second convolution feature extraction block and continuous with the second convolution feature extraction block
Convolution feature extraction block.Convolution feature extraction block includes L convolution module, and it includes volume in any one convolution module that L, which is greater than 0,
Lamination, BN layers and excitation layer, it is specific as shown in Figure 6.
Further, after training identification model, the pisiform in carpal area is identified using identification model, including following
Step, as shown in Figure 7:
The stone age piece of carpal area is handled through N number of continuous convolution feature extraction block, obtains carpal area by step S701
Characteristics of image.
The characteristics of image of carpal area is inputted full articulamentum, exports the recognition result of pisiform by step S702.
The recognition result of the pisiform of full articulamentum output includes pisiform and without two kinds of pisiform, then further root
The stone age of user to be detected is determined according to the recognition result of pisiform.
Since in carpal area, pisiform goes out to represent specific stage stone age, therefore the embodiment of the present application
In, the pisiform in carpal area is individually identified using identification model, user to be detected is determined according to the recognition result of pisiform
Stone age, do not need human subjective according to stone age piece and judge the stone age, thus on the one hand improve detection the stone age precision, another party
Face improves the efficiency of detection stone age.Secondly, pisiform can be Chong Die with ostriquetrum after growing, therefore ossify in detection of carpal bone region
When the quantity at center determines the stone age, pisiform and ostriquetrum may be identified as one piece of bone, and therefore, identification pisiform can be right
It determines that the stone age is supplemented by the quantity of ossification centre in detection of carpal bone region, improves the precision of detection stone age.
Based on the same technical idea, the embodiment of the present application provides a kind of device for detecting the stone age, as shown in figure 8, should
Device 800 includes:
Module 801 is obtained, for obtaining the stone age piece of user to be detected, and adjusts each bone in the stone age piece to base
Level is set；
Divide module 802, for determining carpal area from stone age piece；
Identification module 803, for identifying the pisiform in the carpal area using identification model, the identification model is
Using mark in advance pisiform carpal area stone age piece as training sample, after being trained to depth residual error network determine；
Detection module 804, for determining the stone age of the user to be detected according to the recognition result of the pisiform.
Optionally, the segmentation module 802 is specifically used for:
The coordinate of the corresponding key point of carpal area in the stone age piece, the parted pattern are determined using parted pattern is
Using mark in advance key point coordinate multiple stone age pieces as training sample, after being trained to depth residual error network determine；
The carpal area is determined according to the coordinate of the corresponding key point of the carpal area.
Optionally, the identification module 803 is specifically used for:
The stone age piece of the carpal area is handled through N number of continuous convolution feature extraction block, obtains the carpal area
Characteristics of image, N is greater than 0, and the convolution feature extraction block includes L convolution module, and L is greater than 0, in any one convolution module
Including convolutional layer, BN layers and excitation layer；For the continuous first convolution feature of any two in N number of convolution feature extraction block
Extract block and the second convolution feature extraction block, the second characteristics of image and described first of the second convolution feature extraction block output
Convolution feature extraction block output the first characteristics of image be added after as third product feature extraction block input or it is N number of continuously
The output of convolution feature extraction block；The third convolution feature extraction block be after the second convolution feature extraction block and
With the continuous convolution feature extraction block of the second convolution feature extraction block；
The characteristics of image of the carpal area is inputted into full articulamentum, exports the recognition result of the pisiform.
Optionally, the acquisition module 801 is specifically used for:
Obtain the coordinate of preset reference point；
The coordinate of key point in the stone age piece is determined using adjustment model, the adjustment model is to mark key in advance
Multiple stone age pieces of the coordinate of point are training sample, are determined after being trained to depth residual error network, the seat of the key point
The coordinate of mark and the preset reference point is the coordinate in the same coordinate system；
The current of each bone of stone age piece is determined according to the coordinate of the coordinate of the preset reference point and the key point
Corresponding relationship between position and base position；
Each bone in the stone age piece is adjusted to base position according to the corresponding relationship.
Optionally, the detection module 804 is specifically used for:
When including pisiform in determining the carpal area, determine that the stone age of the user to be detected is first default
Value；
When not including pisiform in determining the carpal area, determine that the stone age of the user to be detected is second default
Value.
Based on the same technical idea, the embodiment of the present application provides a kind of equipment for detecting the stone age, as shown in figure 9, packet
At least one processor 901 is included, and the memory 902 connecting at least one processor, is not limited in the embodiment of the present application
Specific connection medium between processor 901 and memory 902 passes through bus between processor 901 and memory 902 in Fig. 9
For connection.Bus can be divided into address bus, data/address bus, control bus etc..
In the embodiment of the present application, memory 902 is stored with the instruction that can be executed by least one processor 901, at least
The instruction that one processor 901 is stored by executing memory 902 can execute included in the method for detection stone age above-mentioned
The step of.
Wherein, processor 901 is the control centre for detecting the equipment of stone age, can use various interfaces and connection inspection
The various pieces for surveying the equipment of stone age, are stored in by running or executing the instruction being stored in memory 902 and calling
Data in reservoir 902, to realize the detection stone age.Optionally, processor 901 may include one or more processing units, place
Reason device 901 can integrate application processor and modem processor, wherein the main processing operation system of application processor, user
Interface and application program etc., modem processor mainly handle wireless communication.It is understood that above-mentioned modulation /demodulation processing
Device can not also be integrated into processor 901.In some embodiments, processor 901 and memory 902 can be in same chips
Upper realization, in some embodiments, they can also be realized respectively on independent chip.
Processor 901 can be general processor, such as central processing unit (CPU), digital signal processor, dedicated integrated
Circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array or other can
Perhaps transistor logic, discrete hardware components may be implemented or execute the application implementation for programmed logic device, discrete gate
Each method, step and logic diagram disclosed in example.General processor can be microprocessor or any conventional processor
Deng.The step of method in conjunction with disclosed in the embodiment of the present application, can be embodied directly in hardware processor and execute completion, Huo Zheyong
Hardware and software module combination in processor execute completion.
Memory 902 is used as a kind of non-volatile computer readable storage medium storing program for executing, can be used for storing non-volatile software journey
Sequence, non-volatile computer executable program and module.Memory 902 may include the storage medium of at least one type,
It such as may include flash memory, hard disk, multimedia card, card-type memory, random access storage device (Random Access
Memory, RAM), static random-access memory (Static Random Access Memory, SRAM), may be programmed read-only deposit
Reservoir (Programmable Read Only Memory, PROM), read-only memory (Read Only Memory, ROM), band
Electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory,
EEPROM), magnetic storage, disk, CD etc..Memory 902 can be used for carrying or storing have instruction or data
The desired program code of structure type and can by any other medium of computer access, but not limited to this.The application is real
Applying the memory 902 in example can also be circuit or other devices that arbitrarily can be realized store function, for storing program
Instruction and/or data.
Based on the same inventive concept, the embodiment of the present application also provides a kind of computer-readable medium, being stored with can be by
The computer program that the equipment for detecting the stone age executes, when described program is run in the equipment for detecting the stone age, so that the inspection
The step of equipment for surveying the stone age executes the method for detection stone age.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the present invention
Form.It is deposited moreover, the present invention can be used to can be used in the computer that one or more wherein includes computer usable program code
The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (12)
1. a kind of method for detecting the stone age characterized by comprising
The stone age piece of user to be detected is obtained, and adjusts each bone in the stone age piece to base position；
Carpal area is determined from the stone age piece；
Pisiform in the carpal area is identified using identification model, the identification model is the wrist to mark pisiform in advance
The stone age piece in bone region is training sample, is determined after being trained to depth residual error network；
The stone age of the user to be detected is determined according to the recognition result of the pisiform.
2. the method as described in claim 1, which is characterized in that described to determine carpal area from the stone age piece, comprising:
Determine that the coordinate of the corresponding key point of carpal area in the stone age piece, the parted pattern are with pre- using parted pattern
Multiple the stone age pieces for first marking the coordinate of key point are training sample, are determined after being trained to depth residual error network；
The carpal area is determined according to the coordinate of the corresponding key point of the carpal area.
3. the method as described in claim 1, which is characterized in that the pea identified using identification model in the carpal area
Beans bone, comprising:
The stone age piece of the carpal area is handled through N number of continuous convolution feature extraction block, obtains the figure of the carpal area
As feature, N is greater than 0, and the convolution feature extraction block includes L convolution module, and L is greater than 0, includes in any one convolution module
Convolutional layer, BN layers and excitation layer；For the continuous first convolution feature extraction of any two in N number of convolution feature extraction block
Block and the second convolution feature extraction block, the second characteristics of image of the second convolution feature extraction block output and first convolution
Input or N number of continuous convolution after the first characteristics of image addition of feature extraction block output as third product feature extraction block
The output of feature extraction block；The third convolution feature extraction block for after the second convolution feature extraction block and with institute
State the second continuous convolution feature extraction block of convolution feature extraction block；
The characteristics of image of the carpal area is inputted into full articulamentum, exports the recognition result of the pisiform.
4. the method as described in claim 1, which is characterized in that each bone adjusted in the stone age piece to benchmark position
It sets, comprising:
Obtain the coordinate of preset reference point；
The coordinate of key point in the stone age piece is determined using adjustment model, the adjustment model is to mark key point in advance
Multiple stone age pieces of coordinate be training sample, after being trained to depth residual error network determine, the coordinate of the key point and
The coordinate of the preset reference point is the coordinate in the same coordinate system；
The current location of each bone of stone age piece is determined according to the coordinate of the coordinate of the preset reference point and the key point
Corresponding relationship between base position；
Each bone in the stone age piece is adjusted to base position according to the corresponding relationship.
5. the method as described in Claims 1-4 is any, which is characterized in that the recognition result according to the pisiform is true
The stone age of the fixed user to be detected, comprising:
When including pisiform in determining the carpal area, determine that the stone age of the user to be detected is the first preset value；
When not including pisiform in determining the carpal area, determine that the stone age of the user to be detected is the second preset value.
6. a kind of device for detecting the stone age characterized by comprising
Module is obtained, for obtaining the stone age piece of user to be detected, and adjusts each bone in the stone age piece to base position；
Divide module, for determining carpal area from the stone age piece；
Identification module, for identifying the pisiform in the carpal area using identification model, the identification model is with preparatory
The stone age piece for marking the carpal area of pisiform is training sample, is determined after being trained to depth residual error network；
Detection module, for determining the stone age of the user to be detected according to the recognition result of the pisiform.
7. device as claimed in claim 6, which is characterized in that the segmentation module is specifically used for:
Determine that the coordinate of the corresponding key point of carpal area in the stone age piece, the parted pattern are with pre- using parted pattern
Multiple the stone age pieces for first marking the coordinate of key point are training sample, are determined after being trained to depth residual error network；
The carpal area is determined according to the coordinate of the corresponding key point of the carpal area.
8. device as claimed in claim 6, which is characterized in that the identification module is specifically used for:
The stone age piece of the carpal area is handled through N number of continuous convolution feature extraction block, obtains the figure of the carpal area
As feature, N is greater than 0, and the convolution feature extraction block includes L convolution module, and L is greater than 0, includes in any one convolution module
Convolutional layer, BN layers and excitation layer；For the continuous first convolution feature extraction of any two in N number of convolution feature extraction block
Block and the second convolution feature extraction block, the second characteristics of image of the second convolution feature extraction block output and first convolution
Input or N number of continuous convolution after the first characteristics of image addition of feature extraction block output as third product feature extraction block
The output of feature extraction block；The third convolution feature extraction block for after the second convolution feature extraction block and with institute
State the second continuous convolution feature extraction block of convolution feature extraction block；
The characteristics of image of the carpal area is inputted into full articulamentum, exports the recognition result of the pisiform.
9. device as claimed in claim 6, which is characterized in that the acquisition module is specifically used for:
Obtain the coordinate of preset reference point；
The coordinate of key point in the stone age piece is determined using adjustment model, the adjustment model is to mark key point in advance
Multiple stone age pieces of coordinate be training sample, after being trained to depth residual error network determine, the coordinate of the key point and
The coordinate of the preset reference point is the coordinate in the same coordinate system；
The current location of each bone of stone age piece is determined according to the coordinate of the coordinate of the preset reference point and the key point
Corresponding relationship between base position；
Each bone in the stone age piece is adjusted to base position according to the corresponding relationship.
10. the device as described in claim 6 to 9 is any, which is characterized in that the detection module is specifically used for:
When including pisiform in determining the carpal area, determine that the stone age of the user to be detected is the first preset value；
When not including pisiform in determining the carpal area, determine that the stone age of the user to be detected is the second preset value.
11. a kind of equipment for detecting the stone age, which is characterized in that including at least one processor and at least one processor,
In, the storage unit is stored with computer program, when described program is executed by the processor, so that the processor is held
The step of row Claims 1 to 5 any claim the method.
12. a kind of computer-readable medium, which is characterized in that it is stored with the computer journey that can be executed by the equipment for detecting the stone age
Sequence, when described program is run in the equipment for detecting the stone age, so that the equipment perform claim of the detection stone age requires 1~5
The step of any the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811161771.3A CN109285154A (en) | 2018-09-30 | 2018-09-30 | A kind of method and device detecting the stone age |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811161771.3A CN109285154A (en) | 2018-09-30 | 2018-09-30 | A kind of method and device detecting the stone age |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109285154A true CN109285154A (en) | 2019-01-29 |
Family
ID=65176473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811161771.3A Pending CN109285154A (en) | 2018-09-30 | 2018-09-30 | A kind of method and device detecting the stone age |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109285154A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110211674A (en) * | 2019-04-23 | 2019-09-06 | 平安科技（深圳）有限公司 | Stone age test method and relevant device based on machine learning model |
| CN110335670A (en) * | 2019-06-10 | 2019-10-15 | 北京深睿博联科技有限责任公司 | Image processing method and device for the classification of epiphysis grade |
| CN110738654A (en) * | 2019-10-18 | 2020-01-31 | 中国科学技术大学 | Key point extraction and bone age prediction method in hip joint image |
| CN110782450A (en) * | 2019-10-31 | 2020-02-11 | 北京推想科技有限公司 | Hand carpal development grade determining method and related equipment |
| CN111507953A (en) * | 2020-04-13 | 2020-08-07 | 武汉华晨酷神智能科技有限公司 | AI bone age rapid identification method |
| CN112438745A (en) * | 2019-09-04 | 2021-03-05 | 株式会社岛津制作所 | Image analysis method, processing device, measuring device, and learning model creation method |
| CN113647965A (en) * | 2021-08-30 | 2021-11-16 | 北京理工大学 | Forearm bone identification method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107591200A (en) * | 2017-08-25 | 2018-01-16 | 卫宁健康科技集团股份有限公司 | Stone age marker recognition appraisal procedure and system based on deep learning and image group |
| CN107895367A (en) * | 2017-11-14 | 2018-04-10 | 中国科学院深圳先进技术研究院 | A kind of stone age recognition methods, system and electronic equipment |
| CN108056786A (en) * | 2017-12-08 | 2018-05-22 | 浙江大学医学院附属儿童医院 | A kind of stone age detection method and device based on deep learning |
-
2018
- 2018-09-30 CN CN201811161771.3A patent/CN109285154A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107591200A (en) * | 2017-08-25 | 2018-01-16 | 卫宁健康科技集团股份有限公司 | Stone age marker recognition appraisal procedure and system based on deep learning and image group |
| CN107895367A (en) * | 2017-11-14 | 2018-04-10 | 中国科学院深圳先进技术研究院 | A kind of stone age recognition methods, system and electronic equipment |
| CN108056786A (en) * | 2017-12-08 | 2018-05-22 | 浙江大学医学院附属儿童医院 | A kind of stone age detection method and device based on deep learning |
Non-Patent Citations (2)
| Title |
|---|
| KAIMING HE ET AL: "Deep Residual Learning for Image Recognition", 《2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION(CVPR)》 * |
| 崔玮华等: "对七十九名少儿的骨龄分析", 《体育教学与科研》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110211674A (en) * | 2019-04-23 | 2019-09-06 | 平安科技（深圳）有限公司 | Stone age test method and relevant device based on machine learning model |
| CN110211674B (en) * | 2019-04-23 | 2023-06-27 | 平安科技（深圳）有限公司 | Bone age testing method and related equipment based on machine learning model |
| CN110335670A (en) * | 2019-06-10 | 2019-10-15 | 北京深睿博联科技有限责任公司 | Image processing method and device for the classification of epiphysis grade |
| CN112438745A (en) * | 2019-09-04 | 2021-03-05 | 株式会社岛津制作所 | Image analysis method, processing device, measuring device, and learning model creation method |
| CN110738654A (en) * | 2019-10-18 | 2020-01-31 | 中国科学技术大学 | Key point extraction and bone age prediction method in hip joint image |
| CN110738654B (en) * | 2019-10-18 | 2022-07-15 | 中国科学技术大学 | Key point extraction and bone age prediction method in hip joint image |
| CN110782450A (en) * | 2019-10-31 | 2020-02-11 | 北京推想科技有限公司 | Hand carpal development grade determining method and related equipment |
| CN111507953A (en) * | 2020-04-13 | 2020-08-07 | 武汉华晨酷神智能科技有限公司 | AI bone age rapid identification method |
| CN113647965A (en) * | 2021-08-30 | 2021-11-16 | 北京理工大学 | Forearm bone identification method and system |
| CN113647965B (en) * | 2021-08-30 | 2023-11-28 | 北京理工大学 | Forearm bone identification method and forearm bone identification system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109146879A (en) | A kind of method and device detecting the stone age | |
| CN109285154A (en) | A kind of method and device detecting the stone age | |
| CN109377484A (en) | A kind of method and device detecting the stone age | |
| CN109272002A (en) | A kind of classification method and device of stone age piece | |
| CN109255786A (en) | A kind of method and device detecting the stone age | |
| CN108764241A (en) | Divide method, apparatus, computer equipment and the storage medium of near end of thighbone | |
| CN110309706A (en) | Face critical point detection method, apparatus, computer equipment and storage medium | |
| CN109509177B (en) | Method and device for recognizing brain image | |
| CN109447963B (en) | Brain image recognition method and device | |
| CN109815865A (en) | A kind of water level recognition methods and system based on virtual water gauge | |
| CN111275686B (en) | Method and device for generating medical image data for artificial neural network training | |
| CN110533639B (en) | Key point positioning method and device | |
| CN107871316A (en) | A kind of X-ray hand bone interest region extraction method based on deep neural network | |
| CN109523546A (en) | A kind of method and device of Lung neoplasm analysis | |
| CN109816721A (en) | Image position method, device, equipment and storage medium | |
| CN111612756A (en) | Coronary artery specificity calcification detection method and device | |
| CN110400302A (en) | The method and device of lesion information in a kind of determination, display breast image | |
| CN114359248A (en) | Medical image segmentation method and device, storage medium and electronic equipment | |
| CN110992310A (en) | Method and device for determining partition where mediastinal lymph node is located | |
| CN113642581B (en) | Image semantic segmentation method and system based on coding multipath semantic cross network | |
| CN114693671A (en) | Lung nodule semi-automatic segmentation method, device, equipment and medium based on deep learning | |
| CN111815569A (en) | Image segmentation method, device and equipment based on deep learning and storage medium | |
| CN113408595B (en) | Pathological image processing method and device, electronic equipment and readable storage medium | |
| CN111046870B (en) | Finger joint positioning method, device, medium and equipment | |
| CN110390671B (en) | Method and device for detecting mammary gland calcification |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Fu Junfen Inventor after: Wei Zikun Inventor after: Hua Yiwei Inventor after: Ding Zezhen Inventor before: Wei Zikun Inventor before: Hua Yiwei Inventor before: Ding Zezhen |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |