CN109308708A - Image processing method, device and the retina stimulator of low pixel - Google Patents
Image processing method, device and the retina stimulator of low pixel Download PDFInfo
- Publication number
- CN109308708A CN109308708A CN201811047463.8A CN201811047463A CN109308708A CN 109308708 A CN109308708 A CN 109308708A CN 201811047463 A CN201811047463 A CN 201811047463A CN 109308708 A CN109308708 A CN 109308708A
- Authority
- CN
- China
- Prior art keywords
- area
- target object
- target
- image processing
- pixel
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- 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
- G06T2207/30041—Eye; Retina; Ophthalmic
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present disclosure describes a kind of image processing methods of low pixel characterized by comprising obtains initial pictures；Background area and target area including multiple target objects are distinguished from initial pictures, removes background area；The area of each target object is calculated from target area；And the area for removing target object is less than the target object of defined threshold, and the area for retaining target object is greater than or equal to the target object of defined threshold.Thereby, it is possible to the image informations of target object needed for effectively retaining patient to help patient to identify target object.
Description
Technical field
This disclosure relates to bionics techniques field, and in particular to a kind of image processing method of low pixel, device and view
Film stimulator.
Background technique
The formation of normal vision is that external optical signal is converted to vision letter by the photosensory cell on intraocular retina
Number.Visual signal reaches cerebral cortex via Beale's ganglion cells and gangliocyte, to form light sensation.However, suffering from life
Person usually makes because of a variety of retinal diseases such as RP (retinal pigment degeneration), AMD (with old related macular degeneration) etc.
Photosensitive access is obstructed, and leads to vision decline or blinding.
With the research and development of technology, occur repairing above-mentioned view using artificial retina or retina stimulator etc.
The technological means of Omental lesion can allow brain that can receive environmental stimuli signal and obtain improved view by the technology
Feel.Existing retina stimulator system generally comprises the photographic device and video process apparatus of arrangement outside the patient's body, and
It is put into the implant of the eyeball of patient.External photographic device captures initial pictures, and obtained image is converted into vision
Signal, video process apparatus are sent to implant after being handled visual signal, implant is further by these visual signals
It is converted to electrical stimulation signal, to stimulate gangliocyte or the Beale's ganglion cells on retina, to generate light sensation to patient.
It in existing video process apparatus, needs to compress visual signal, so that compressed visual signal energy
The stimulating electrode of enough matched implants.However, when photographic device acquires some more complicated scenes, some lesser objects
Body may will become one or two of pixel after overcompression, be not enough to that patient is allowed to differentiate, and these lesser objects are at
The pixel obtained after reason may become " noise ", interfere the recognition capability of patient.
Summary of the invention
The disclosure in view of the above-mentioned prior art situation and complete, its purpose is to provide one kind effectively to protect
The image information of target object needed for staying patient is to help the image processing method of the low pixel of patient's identification target object, fill
It sets and retina stimulator.
For this purpose, the first aspect of the disclosure provides a kind of image processing method of low pixel characterized by comprising
Obtain initial pictures；Background area and target area including multiple target objects are distinguished from the initial pictures, described in removal
Background area；The area of each target object is calculated from the target area；And remove the face of the target object
Product is less than the target object of defined threshold, and the area for retaining the target object is greater than or equal to the target pair of the defined threshold
As.
In the disclosure, background area and the target area including multiple target objects are distinguished, removal background area retains
Target area, calculate target area in target object area, remain larger than or equal to defined threshold target object.At this
In the case of kind, removal background area can tentatively reduce image pixel, and removal is less than the target object of defined threshold, can be effective
The image information of target object needed for ground effectively retains patient is to help patient to identify target object.
In the image processing method involved in the disclosure, optionally, by the pixel quantity for counting the target object
To calculate the area of the target object.Thereby, it is possible to the area of target object is obtained based on pixel quantity.
In addition, in the image processing method involved in the disclosure, optionally, in the area for calculating the target object
In, choose multiple coordinate points of the perimeter along the target object so that multiple coordinate points connections and formed
Region is fitted the area of the target object, and calculates the area in the region.Thereby, it is possible to the fitting areas based on coordinate points
The area of domain acquisition target object.
In addition, in the image processing method involved in the disclosure, optionally, in the area for calculating the target object
In, limb recognition is carried out for the target object, counts the pixel quantity at the edge of the target object.Thereby, it is possible to bases
Pixel quantity analogy in edge obtains the area of target object.
In addition, in the image processing method involved in the disclosure, can by the area of each target object according to
Drop rises minor sort, selects the intermediate value in putting in order as the defined threshold.Thereby, it is possible to easily obtain regulation threshold
Value.
In addition, being averaged for the area of the target object can be calculated in the image processing method involved in the disclosure
Value, the average value is as the defined threshold.Thereby, it is possible to easily obtain defined threshold.
In addition, the area of the target object can be enabled to be less than regulation in the image processing method involved in the disclosure
The pixel value of the target object of threshold value is set as zero.The target object for being less than defined threshold thereby, it is possible to remove area.
The second aspect of the disclosure provides a kind of image processing apparatus of low pixel characterized by comprising obtains mould
Block is used to obtain initial pictures；Divide module, is used to distinguish background area from the initial pictures and including multiple targets
The target area of object removes the background area；Computing module is used to from the target area calculate each mesh
Mark the area of object；And selecting module, the area for being used to remove the target object are less than the target object of defined threshold,
The area for retaining the target object is greater than or equal to the target object of the defined threshold.
In the disclosure, segmentation module distinguishes background area and the target area including multiple target objects, removes background
Region retains target area, and computing module calculates the area of the target object in target area, and selecting module is remained larger than or waited
In the target object of defined threshold.In this case, removal background area can tentatively reduce image pixel, and removal is less than rule
Determine the target object of threshold value, the image information of target object needed for effectively retaining patient is to help patient to identify mesh
Mark object.
In the image processing apparatus involved in the disclosure, the computing module can be by counting the target object
Pixel quantity calculates the area of the target object.Thereby, it is possible to the area of target object is obtained based on pixel quantity.
In addition, optionally, the computing module is chosen along the mesh in the image processing apparatus involved in the disclosure
Multiple coordinate points of the perimeter of object are marked, so that the region that multiple coordinate points are connected and formed is fitted the target pair
The area of elephant, and calculate the area in the region.Thereby, it is possible to the faces that the fitted area based on coordinate points obtains target object
Product.
In addition, optionally, the target is directed in the computing module in the image processing apparatus involved in the disclosure
Object carries out limb recognition, counts the pixel quantity at the edge of the target object.Thereby, it is possible to the pixel quantities based on edge
The area of analogy acquisition target object.
In addition, the selecting module can be by each target pair in the image processing apparatus involved in the disclosure
The area of elephant selects the intermediate value in putting in order as the defined threshold according to drop minor sort.Thereby, it is possible to easily obtain
Defined threshold.
In addition, optionally, the selecting module calculates the target pair in the image processing apparatus involved in the disclosure
The average value of the area of elephant, the average value is as the defined threshold.Thereby, it is possible to easily obtain defined threshold.
In addition, optionally, the selecting module enables the target object in the image processing apparatus involved in the disclosure
Area be less than the pixel value of target object of defined threshold and be set as zero.Thereby, it is possible to remove area less than defined threshold
Target object.
The third aspect of the disclosure provides a kind of retina stimulator characterized by comprising photographic device is used
In capture video image, and the video image is converted into visual signal；Video process apparatus includes at least above-mentioned
Image processing apparatus described in one, the video process apparatus are connect with the photographic device, and the video process apparatus is used
In by the visual signal carry out handle and be sent to implanted device via transmitting antenna；And the implanted device, it is used for
The received visual signal of institute is converted into pulsed current signal, to provide the pulsed current signal to retina.
According to the disclosure, be capable of providing it is a kind of can effectively retain patient needed for target object image letter
Cease image processing method, device and the retina stimulator of the low pixel to help patient to identify target object.
Detailed description of the invention
Embodiment of the disclosure will be explained in further detail solely by reference to the example of attached drawing now, in which:
Fig. 1 is the structural schematic diagram of retina stimulator involved in the disclosure.
Fig. 2 is the image processing apparatus structural schematic diagram of low pixel involved in the disclosure.
Fig. 3 is the image processing apparatus structural schematic diagram of low pixel involved in the disclosure.
Fig. 4 is the flow diagram of the figure pixel processing method of low pixel involved in the disclosure.
Fig. 5 is the flow diagram of the example of area computation method involved in the disclosure.
Fig. 6 is the flow diagram of the variation 1 of area computation method involved in the disclosure.
Fig. 7 is the flow diagram of the variation 2 of area computation method involved in the disclosure.
Specific embodiment
Hereinafter, explaining the preferred embodiment of the present invention in detail with reference to attached drawing.In the following description, for identical
Component assign identical symbol, the repetitive description thereof will be omitted.Scheme in addition, attached drawing is only schematical, the mutual ruler of component
Very little shape of ratio or component etc. can be with actual difference.
In addition, below the present invention subhead involved in description etc. be not meant to limit the present invention in perhaps model
It encloses, is merely possible to the suggesting effect read.Such subhead can neither be interpreted as the content for dividing article, also not
Content under subhead should be limited only in the range of subhead.
Fig. 1 is the structural schematic diagram of retina stimulator involved in the disclosure.This disclosure relates to retina stimulator 1
It can be adapted for retinopathy and cause to blind, but the pathways for vision such as Beale's ganglion cells, gangliocyte retain intact patient.
Retina stimulator 1 is referred to as " retina stimulator system ", " artificial retina ", " artificial retina ", " artificial sometimes
Retina system ", " artificial retina system " etc..
In some instances, as shown in Figure 1, retina stimulator 1 may include photographic device 10, video process apparatus 20
With implanted device 30.Implanted device 30 can receive visual signal and view-based access control model signal generates pulsed current signal.Wherein, depending on
Feel that signal can be acquired by photographic device 10, and handles and obtain via video process apparatus 20.
In some instances, photographic device 10 can be used for capturing video image, and video image is converted into vision
Signal.For example, photographic device 10 can capture the video image of patient's local environment.
In some instances, photographic device 10 can be equipment with camera function, such as video camera, camera etc..
It, can be by the Camera Design of small volume on (such as being embedded into) glasses in order to facilitate use.
In other examples, patient can also be by wearing the light glasses with camera function as photographic device
10 capture video image.For example, photographic device 10 can be realized with Google glass etc..In addition, photographic device 10 can fill
Fit over intelligent glasses, intelligence is worn, on the intelligent wearable device such as Intelligent bracelet.
In some instances, video process apparatus 20 can be connect with photographic device 10.Photographic device 10 can be by having
Line connection is wirelessly connected and video process apparatus 20.
In some instances, wired connection can be data line connection.In addition, in some instances, wireless connection can be with
It is bluetooth connection, WiFi connection, infrared connection, NFC connection or radio frequency connection etc..
In other examples, it is external that video process apparatus 20 and photographic device 10 can be only fitted to patient.For example, patient
Photographic device 10 can be configured on glasses.Patient can also configure photographic device 10 in such as headwear, hair band or brooch
Etc. on wearable accessories.In addition, patient can configure video process apparatus 20 in waist, patient can also be handled video
Device 20 configures at such as positions such as arm, leg.The example of the disclosure is without being limited thereto, for example, patient can also will be at video
Reason device 20 is placed in for example portable handbag or knapsack.
In some instances, video process apparatus 20 can receive the visual signal of the generation of photographic device 10.Video processing
Device 20 handles visual signal, and is sent to implanted device 30 via transmitting antenna.
In addition, in some instances, video process apparatus 20 may include the image processing apparatus of low pixel.Low pixel
The dependent module and function of image processing apparatus are described in detail later.
In some instances, implanted device 30 can be used for the received visual signal of institute being converted into pulsed current signal,
To provide pulsed current signal to retina.
In some instances, implanted device 30 may include the stimulating electrode of specified quantity.Stimulating electrode is (sometimes referred to as
" electrode ") electrical stimulation signal can be generated according to visual signal.Specifically, implanted device 30 can receive visual signal, and
And stimulating electrode can by received visual signal be converted into for example two-way arteries and veins of pulsed current signal as electrical stimulation signal
Current signal is rushed, so that the gangliocyte or Beale's ganglion cells to the tissue such as retina of retina provide Bipolar pulse current
Signal generates light sensation.In addition, implanted device 30 can be implanted into human body such as eyeball.
Fig. 2 is the image processing apparatus structural schematic diagram of low pixel involved in the disclosure.Low picture involved in the disclosure
The image processing apparatus 200 (can be referred to as image processing apparatus 200) of element can be used for retina stimulator 1 as at image
The functional module of reason.The image processing apparatus 200 of low pixel may include in video process apparatus 20.
In some instances, as shown in Fig. 2, image processing apparatus 200 may include obtaining module 211, segmentation module
212, computing module 213 and selecting module 214.Segmentation module 212 can distinguish the back for obtaining the initial pictures that module 211 obtains
Scene area and target area, and remove background area and retain target area.Computing module 213 can calculate the mesh in target area
Mark object area, selecting module 214 can remain larger than or equal to defined threshold target object.
In some instances, obtaining module 211 can be used for obtaining initial pictures.Initial pictures can be color image,
Initial pictures can also be gray level image.Specifically, obtaining the visual signal that module 211 can be exported based on photographic device 10
Obtain initial pictures.Wherein, the pixel of initial pictures can be determined by the pixel of the pick-up lens (not shown) of photographic device 10.
For example, the pixel of pick-up lens can be for such as 300,000,500,000,1,000,000,2,000,000,5,000,000,12,000,000.Initial pictures
Pixel quantity is correspondingly also possible to and the pixel of shots match such as 300,000,500,000,1,000,000,2,000,000,5,000,000,12,000,000
Deng.
In some instances, segmentation module 212 can be used for distinguishing background area from initial pictures and including multiple target
The target area of object removes background area.
In some instances, initial pictures usually may include background area and target area.Wherein, target area can be with
It is the region comprising patient's information needed (such as object or barrier).Background area can be in initial pictures target area with
Outer region.For example, the initial pictures obtained based on environment locating for patient, wherein the region where object or barrier can
To be target area, object or the unexpected region of barrier can be background area.
In some instances, it can wrap in target area containing one, two or more target objects.For example, object
Body or barrier can be the target object of target area.The quantity of target object is the quantity of object or barrier.
It in some instances, include back in the pixel of initial pictures due to a part that background area is initial pictures
The pixel of scene area.In this case, also the pixel of background area is also handled when handling initial pictures.
Generally, due to be implanted into the limitation of space and design technology, the stimulation electricity of the implant of patient's eyeball is set
Ultimate ratio is relatively limited such as 60,120 electrodes, 256 electrodes, and stimulating electrode reception information capability is very limited, if directly
It connects and corresponds to the limited stimulating electrode that implants with the 10 captured image information of photographic device of retina stimulator 1, then can
A large amount of information loss is caused, image fault is caused.In addition, the background area of the initial pictures captured by photographic device 10 occupies
A part of initial pictures, the specific gravity of the pixel of target area needed for making patient in initial pictures reduce, thus cause
The image information of target area cannot pass to patient well.In this case, in the present embodiment, relative to background
Region, the information that target area is included are more useful information for the patient, can be removed by segmentation module 212
Background area.The target pair for being greater than or equal to defined threshold thereby, it is possible to improve area in the target area in subsequent image
As specific gravity in the picture, optimize the subsequent processing to target area, preferably help patient identifies target object.
In some instances, segmentation module 212 can by threshold value comparison method distinguish initial pictures in background area and
Target area.Specifically, threshold value comparison method, using preset threshold and initial pictures, can be distinguished by the way that preset threshold is arranged
Background area and target area.
In some instances, preset threshold can be the average pixel value of initial pictures.Wherein, average pixel value can be
The average pixel value of whole pixels of initial pictures.The example of the disclosure is without being limited thereto, and preset threshold can also be initial pictures
Average gray value.
In some instances, average pixel value can also be the average pixel value of the partial pixel of initial pictures.Part picture
Element can be obtained based on the mode of the pixel value of initial pictures.Specifically, can pixel value to the pixel of initial pictures into
Row arrangement, obtains the mode of the pixel value of initial pictures.The pixel that pixel value is equal to mode is selected, partial pixel is obtained.This public affairs
The example opened is without being limited thereto, and partial pixel can be obtained based on preset ratio.For example, can arrange the pixel of initial pictures
Column, by the pixel for obtaining the initial pictures of preset ratio that puts in order as partial pixel.Wherein, preset ratio can be part
The accounting of pixel and initial pictures whole pixel.Preset ratio can be such as 50%, 60% or 80%.
In some instances, putting in order can be ascending or descending order, can also according to pixel position from left to right, from
Top to bottm arrangement.But it puts in order and is not limited to above situation.
In some instances, preset threshold can be the intermediate value of the pixel value of initial pictures.Specifically, can will be initial
The pixel of image carries out ascending or descending order arrangement, the intermediate value of the pixel of initial pictures is determined, to obtain preset threshold.
In some instances, preset threshold can be obtained based on each region of initial pictures.Specifically, can will be first
Beginning image is divided into multiple regions, calculates the average pixel value of each region, determines the target area in multiple regions and mesh
Mark the average pixel value in region.Wherein, the difference of the average pixel value of target area and other regions is maximum.It can be by target area
Preset threshold of the average pixel value in domain as initial pictures.
In some instances, based on above-mentioned preset threshold, preset threshold and initial pictures can be compared by dividing module 212
Each pixel pixel value, by initial pictures be divided into more than or equal to preset threshold first area and be less than preset threshold
Second area.Usual background area is greater than target area, therefore, can compare first area and second area, select first
It is background area that region is biggish in region and second area.
In some instances, segmentation module 212 can also set zero for the pixel value of the pixel of background area.Usual picture
Plain value is set as zero and can be considered as pixel being not present, that is, background area can be removed by dividing module 212.Background is removed as a result,
Region can reduce the complexity of subsequent image processing.
In some instances, computing module 213 can be used for calculating the area of each target object from target area.Mesh
Mark region may include one, two or more target objects.For example, object or barrier can be the mesh of target area
Mark object.The quantity of target object is the quantity of object or barrier.
In some instances, computing module 213 can calculate target object by counting the pixel quantity of target object
Area.Specifically, different target objects generally takes up, area is not also identical, and big target object is usually than small target
The area that object occupies is big.In addition, the area of pixel quantity and target object is proportional, and in this case, big mesh
It is more to mark pixel quantity of the pixel quantity of object also than small target object.The pixel number of different target objects can be counted
The pixel quantity of each target object, is analogized to the area of each target object by amount.It can be obtained according to pixel quantity as a result,
The area of target object.
In some instances, computing module 213 can choose multiple coordinate points of the perimeter along target object, with
Make multiple coordinate points connections and the area of the region fit object object of formation, and the area of zoning.Wherein, edge week
Enclose may include edge and adjacent edges.That is, the coordinate points chosen can be the point on the edge of target object, it is also possible to
The adjacent edges of target object.
In some instances, the number of the coordinate points of selection can be such as three, four and four with last.Example
Such as, the number of the coordinate points of selection can be three, and the region that three coordinate points connections are formed can fit one closely
Like the delta-shaped region of target object, the area of delta-shaped region is calculated.The area of delta-shaped region can approximation reflect mesh
Mark the area of object.
In some instances, the number of the coordinate points of selection can be four, the area that four coordinate points connections are formed
Domain can fit the quadrilateral area of an approximate target object, calculate the area of quadrilateral area.The face of quadrilateral area
Product approximate can reflect the area of target object.
In some instances, the number of coordinate points can be sat with the marginal point number of approximate target object, in this case
Punctuate connects the region that can fit an approximate target object, calculates the area in the region, the area in the region can be close
Seemingly reflect the area of target object.
In this case, the number of coordinate points is more, fits the region come closer to target object.That is, coordinate
The number of point is more, fits the area of the area closer to target object in the region come.As a result, according to fit come region
The area of target object can be obtained.
In some instances, computing module 213 can carry out limb recognition for target object, count the side of target object
The pixel quantity of edge.Edge can also claim " profile " sometimes.
In some instances, different target objects generally takes up that area is not also identical, and big target object is usually than small
The area that occupies of target object it is big, the profile of corresponding big target object is bigger than the profile of small target object.In addition, picture
The profile of prime number amount and target object is proportional, in such a case, it is possible to by the pixel of the profile of each target object
Quantity analogizes to the area of each target object, is capable of the face of analogy acquisition target object according to the pixel quantity of profile as a result,
Product.
In addition, in some instances, the area that selecting module 214 can be used for removing target object is less than defined threshold
Target object, the area for retaining target object are greater than or equal to the target object of defined threshold.
In some instances, it is specified that threshold value can be obtained according to the area of target object.Selecting module 214 can will be each
The area of target object selects the intermediate value in putting in order as defined threshold according to drop minor sort (or ascending order).
In some instances, it is specified that threshold value can be obtained according to the average value of the area of target object.Selecting module 214 can
To calculate the average value of the area of target object, using average value as defined threshold.Wherein, the average value of the area of target object
It can be the average value of the area of the target object in target area.
In some instances, it is specified that threshold value can be obtained according to the average value of the area of partial target object.It wherein, will be each
The area of a target object chooses area arrangement several target objects before examination as partial target object according to drop minor sort,
The area of calculating section target object, and the average value of the area of calculating section target object.Using the average value of the area as
Defined threshold.
In addition, in some instances, selecting module 214 can enable the area of target object less than the target pair of defined threshold
The pixel value of elephant is set as zero.The target object for being less than defined threshold thereby, it is possible to remove the area of target object.
In some instances, the quantity of stimulating electrode may correspond to the pixel of compressed image.Since setting is being suffered from
The stimulating electrode of the implant of person's eyeball is limited, and reception information capability is limited, therefore, the image information that stimulating electrode can transmit
It is limited.In such a case, it is possible to select the target object that patient is more required in target object by defined threshold.
In some instances, when the quantity of target object (such as object or barrier) includes two or more quantity
When, relative to the lesser target object of area, patient may the more biggish target object of space required information.In this feelings
Under condition, the target object for being less than defined threshold can be removed.That is, the pixel of the target object of defined threshold can will be less than
Value is set as zero.
In some instances, when patient is located near the biggish target object of area, initial pictures and initial pictures
In target area can change, image processing apparatus 200 can carry out the initial pictures after variation at above-mentioned correlation
Reason.For example, the lesser target object of area before variation is likely to become the biggish mesh of area in the initial pictures after variation
Object is marked, patient may see the small target object before above-mentioned initial pictures variation at this time.
In this case, removal background area can tentatively reduce image pixel, and removal is less than the target of defined threshold
Object can further decrease image pixel, can optimize the subsequent processing to target area, and preferably help patient identifies mesh
Mark object.
Here, each unit of above-mentioned mentioned image processing apparatus 200 includes obtaining module 211, segmentation module
212, the function of computing module 213 and selecting module 214 can be realized by the image processing apparatus 200 of following Fig. 3.With
Under explain in detail.
Fig. 3 is the image processing apparatus structural schematic diagram of low pixel involved in the disclosure.In some instances, such as Fig. 3
Shown, which may include processor 221, memory 222 and communication interface 223.
In some instances, processor 221 can be used for carrying out control pipe to the movement that image processing apparatus 200 executes
Reason.For example, processor 221 can be used to implement the function of each unit of above-mentioned image processing apparatus 200.In addition, processing
Device 221 can be also used for that image processing apparatus 200 is supported to execute the step S110 to step S140 in Fig. 4 and/or for herein
Other processes of described technology.
In some instances, processor 221 can be central processing unit (Central Processing Unit, CPU),
General processor, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application-Specific Integrated Circuit, ASIC), field programmable gate array (Field
Programmable Gate Array, FPGA) or other programmable logic device, transistor logic, hardware component or
Person's any combination thereof.It, which may be implemented or executes, combines various illustrative logic blocks, module and electricity described in the disclosure
Road.Processor 221 is also possible to realize the combination of computing function, such as combines comprising one or more microprocessors, DSP and micro-
Combination of processor etc..
In some instances, communication interface 223 can be used for supporting image processing apparatus 200 with other equipment (for example, taking the photograph
As device 10) communication.
In addition, in some instances, communication interface 223 can be communication interface, transceiver, transmission circuit etc..Wherein, lead to
Letter interface 223 is to be referred to as, and may include one or more interfaces.
In some instances, memory 222 can be used for storing the program code and data of image processing apparatus 200.
In addition, in some instances, which can also include communication bus 224.Communication bus 224
It can be Peripheral Component Interconnect standard (Peripheral Component Interconnect, abbreviation PCI) bus or extension work
Industry normal structure (Extended Industry Standard Architecture, abbreviation EISA) bus etc..Communication bus
224 are further divided into address bus, data/address bus, control bus etc..Communication bus 224 can have one or more.For convenient for
It indicates, is only indicated with a line in Fig. 3, it is not intended that an only bus or a type of bus.
It is above-mentioned be this disclosure relates to image processing apparatus 200, below with reference to flow chart detailed description this disclosure relates to
Image processing method.Each step of image processing method can correspond to each unit of image processing apparatus 200.
Fig. 4 is the flow diagram of the figure pixel processing method of low pixel involved in the disclosure.Fig. 5 is involved by the disclosure
And area computation method example flow diagram.Fig. 6 is the variation 1 of area computation method involved in the disclosure
Flow diagram.Fig. 7 is the flow diagram of the variation 2 of area computation method involved in the disclosure.
In some instances, the figure pixel processing method of low pixel can respectively correspond the image procossing dress of above-mentioned low pixel
The method for setting the realization function of 200 each module.
In some instances, as shown in figure 4, the image processing method of low pixel includes obtaining initial pictures (step
S110)。
In step s 110, the visual signal that initial pictures can be exported based on photographic device 10 obtains.In addition, initial graph
The pixel of picture can be determined by the pixel of the pick-up lens (not shown) of photographic device 10.Initial pictures can be color image,
Initial pictures can also be gray level image.
In some instances, as shown in figure 4, the image processing method of low pixel can also include distinguishing to carry on the back from initial pictures
Scene area and target area including multiple target objects remove background area (step S120).
In the step s 120, initial pictures usually may include background area and target area.Wherein, target area can be with
It is the region comprising patient's information needed (such as object or barrier).Background area can be in initial pictures target area with
Outer region.
In some instances, the background area and target area in initial pictures can be distinguished by threshold value comparison method.Tool
For body, threshold value comparison method, using preset threshold and initial pictures, can distinguish background area and mesh by the way that preset threshold is arranged
Mark region.The method for obtaining preset threshold can be in the method in the above-mentioned segmentation module 212 of analogy.
In some instances, the pixel value that each pixel of preset threshold and initial pictures can be compared, by initial pictures
It is divided into the second area more than or equal to the first area of preset threshold and less than preset threshold.Usual background area is greater than mesh
Therefore mark region can compare first area and second area, select region in first area and second area biggish for back
Scene area.
In some instances, after distinguishing background area, zero can be set by the pixel value of the pixel of background area, as
Plain value is set as zero and can be considered as pixel being not present.Background area is removed as a result, can reduce the complexity of subsequent image processing
Degree.
In some instances, as shown in figure 4, the image processing method of low pixel can also include calculating from target area
The area (step S130) of each target object.
In step s 130, target area may include one, two or more target objects.For example, object or
Barrier can be the target object of target area.The quantity of target object is the quantity of object or barrier.
In some instances, as shown in figure 5, step S130 may include the pixel quantity of statistics target object to calculate mesh
Mark the area (step S1310) of object.
In step S1310, big target object is usually bigger than the area that small target object occupies.In addition, pixel number
Amount is proportional with the area of target object, and in this case, the pixel quantity of big target object is also than small target
The pixel quantity of object is more.The pixel quantity that different target objects can be counted, by the pixel quantity class of each target object
Than the area for each target object.The area of target object can be obtained according to pixel quantity as a result,.
In other examples, as shown in fig. 6, step S130 may include the perimeter chosen along target object
Multiple coordinate points (step S1320).
In step S1320, perimeter may include edge and adjacent edges.That is, the coordinate points chosen can be
Point on the edge of target object, is also possible to the adjacent edges of target object.The number of the coordinate points of selection can be for example
Three, four and four with last.
In some instances, as shown in fig. 6, step S130 may include the region fitting of multiple coordinate points connections and formation
The area (step S1321) of target object.
In step S1321, the number of the coordinate points of selection can be three, the area that three coordinate points connections are formed
Domain can fit the delta-shaped region of an approximate target object, the area of delta-shaped region can approximation reflect target pair
The area of elephant.
In some instances, the number of the coordinate points of selection can be four, the area that four coordinate points connections are formed
Domain can fit the quadrilateral area of an approximate target object.The area of quadrilateral area can approximation reflect target pair
The area of elephant.
In this case, the number of coordinate points is more, fits the region come closer to target object.That is, coordinate
The number of point is more, fits the area of the area closer to target object in the region come.
In some instances, as shown in fig. 6, step S130 may include the area (step S1322) of zoning.
In step S1322, according to fit come region shape calculate corresponding region area.For example, fitting
The region come is triangle, can calculate region area with area formula.
In addition, in some instances, as shown in fig. 7, step S130 may include carrying out limb recognition for target object
(step S1330).
In step S1330, big target object is usually bigger than the area that small target object occupies, corresponding big mesh
The profile for marking object is bigger than the profile of small target object.In addition, the profile of pixel quantity and target object is proportional,
In this case, the pixel quantity of the profile of each target object can be analogized to the area of each target object.
In some instances, the profile of target object can be identified by the method for edge detection.
In some instances, as shown in fig. 7, step S130 can also include the pixel quantity at the edge of statistics target object
(step S1331).It is capable of the area of analogy acquisition target object according to the pixel quantity of profile as a result,.
In some instances, as shown in figure 4, the image processing method of low pixel can also include the face of removal target object
Product is less than the target object of defined threshold, and the area for retaining target object is greater than or equal to the target object (step of defined threshold
S140)。
In step S140, the area of each target object can be put in order according to dropping or rising minor sort, selection
Intermediate value as defined threshold.
In addition, the average value of the area of target object can be calculated in step S140, using average value as regulation threshold
Value.Wherein, the average value of the area of target object can be the average value of the area of the target object in target area.
In some instances, it can choose area by the area of each target object according to drop minor sort and arrange before examination
Several target objects are as partial target object, the area of calculating section target object, and the area of calculating section target object
Average value.Using the average value of the area as defined threshold.
In addition, the pixel value for the target object that step S140 can enable the area of target object be less than defined threshold is set as
Zero.
In the disclosure, background area and the target area including multiple target objects are distinguished, removal background area retains
Target area, calculate target area in target object area, remain larger than or equal to defined threshold target object.At this
In the case of kind, removal background area can tentatively reduce image pixel, and removal is less than the target object of defined threshold, can optimize
The subsequent processing to target area, to help patient to identify target object.
Although being illustrated in conjunction with the accompanying drawings and embodiments to the present invention above, it will be appreciated that above description
The invention is not limited in any way.Those skilled in the art without departing from the true spirit and scope of the present invention may be used
To deform and change to the present invention as needed, these deformations and variation are within the scope of the present invention.
Claims (15)
1. a kind of image processing method of low pixel, which is characterized in that
Include:
Obtain initial pictures；
Background area and target area including multiple target objects are distinguished from the initial pictures, removes the background area；
The area of each target object is calculated from the target area；And
Remove the target object area be less than defined threshold target object, the area for retaining the target object be greater than or
Equal to the target object of the defined threshold.
2. image processing method according to claim 1, which is characterized in that
The area of the target object is calculated by counting the pixel quantity of the target object.
3. image processing method according to claim 1, which is characterized in that
In the area for calculating the target object, multiple coordinate points of the perimeter along the target object are chosen, with
The region for making multiple coordinate points connections and being formed is fitted the area of the target object, and calculates the face in the region
Product.
4. image processing method according to claim 1, which is characterized in that
In the area for calculating the target object, limb recognition is carried out for the target object, counts the target object
Edge pixel quantity.
5. image processing method according to claim 1, which is characterized in that
By the area of each target object according to dropping or rising minor sort, select the intermediate value in putting in order as the regulation
Threshold value.
6. image processing method according to claim 1, which is characterized in that
The average value of the area of the target object is calculated, the average value is as the defined threshold.
7. image processing method according to claim 1, which is characterized in that
The pixel value for the target object for enabling the area of the target object be less than defined threshold is set as zero.
8. a kind of image processing apparatus of low pixel, which is characterized in that
Include:
Module is obtained, is used to obtain initial pictures；
Divide module, is used to distinguish background area and the target area including multiple target objects from the initial pictures, go
Except the background area；
Computing module is used to calculate the area of each target object from the target area；And
Selecting module, the area for being used to remove the target object are less than the target object of defined threshold, retain the target
The area of object is greater than or equal to the target object of the defined threshold.
9. image processing apparatus according to claim 8, which is characterized in that
The computing module calculates the area of the target object by counting the pixel quantity of the target object.
10. image processing apparatus according to claim 8, which is characterized in that
The computing module chooses multiple coordinate points of the perimeter along the target object, so that multiple coordinate points
The region for connecting and being formed is fitted the area of the target object, and calculates the area in the region.
11. image processing apparatus according to claim 8, which is characterized in that
Limb recognition is carried out for the target object in the computing module, counts the pixel number at the edge of the target object
Amount.
12. image processing apparatus according to claim 8, which is characterized in that
The selecting module by the area of each target object according to drop minor sort, select intermediate value in putting in order as
The defined threshold.
13. image processing apparatus according to claim 8, which is characterized in that
The selecting module calculates the average value of the area of the target object, and the average value is as the defined threshold.
14. image processing apparatus according to claim 8, which is characterized in that
The pixel value for the target object that the selecting module enables the area of the target object be less than defined threshold is set as zero.
15. a kind of retina stimulator, which is characterized in that
Include:
Photographic device is used to capture video image, and the video image is converted into visual signal；
Video process apparatus, includes at least image processing apparatus described in any one of claim 8 to 14, at the video
Reason device is connect with the photographic device, and the video process apparatus is used to the visual signal be carried out to processing and via transmitting
Antenna is sent to implanted device；And
The implanted device is used to the received visual signal of institute being converted into pulsed current signal, to send out retina
Put the pulsed current signal.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010048619.5A CN111311625A (en) | 2018-09-09 | 2018-09-09 | Image processing method and image processing apparatus |
| CN201811047463.8A CN109308708B (en) | 2018-09-09 | 2018-09-09 | Low-pixel image processing method and device and retina stimulator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811047463.8A CN109308708B (en) | 2018-09-09 | 2018-09-09 | Low-pixel image processing method and device and retina stimulator |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010048619.5A Division CN111311625A (en) | 2018-09-09 | 2018-09-09 | Image processing method and image processing apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109308708A true CN109308708A (en) | 2019-02-05 |
| CN109308708B CN109308708B (en) | 2020-04-03 |
Family
ID=65225011
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811047463.8A Active CN109308708B (en) | 2018-09-09 | 2018-09-09 | Low-pixel image processing method and device and retina stimulator |
| CN202010048619.5A Pending CN111311625A (en) | 2018-09-09 | 2018-09-09 | Image processing method and image processing apparatus |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010048619.5A Pending CN111311625A (en) | 2018-09-09 | 2018-09-09 | Image processing method and image processing apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN109308708B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643373A (en) * | 2019-12-31 | 2021-11-12 | 深圳硅基仿生科技有限公司 | Retina stimulator with bar vision detection |
| CN116929311A (en) * | 2023-09-19 | 2023-10-24 | 中铁第一勘察设计院集团有限公司 | Section deformation monitoring method, device and system for zoom imaging and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070237408A1 (en) * | 2006-04-05 | 2007-10-11 | Kabushiki Kaisha Toshiba | Image processing apparatus and image processing method |
| CN101826228A (en) * | 2010-05-14 | 2010-09-08 | 上海理工大学 | Detection method of bus passenger moving objects based on background estimation |
| CN104599502A (en) * | 2015-02-13 | 2015-05-06 | 重庆邮电大学 | Method for traffic flow statistics based on video monitoring |
| CN104748684A (en) * | 2015-04-13 | 2015-07-01 | 北方工业大学 | Visual detection method and device for crankshaft shoulder back chipping |
| CN105136070A (en) * | 2015-08-21 | 2015-12-09 | 上海植物园 | Method for utilizing electronic device to measure plant leaf area |
| CN106390285A (en) * | 2016-09-30 | 2017-02-15 | 深圳硅基仿生科技有限公司 | Charge compensation circuit, charge compensation method and artificial retina system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107657639A (en) * | 2017-08-09 | 2018-02-02 | 武汉高德智感科技有限公司 | A kind of method and apparatus of quickly positioning target |
| CN108230341B (en) * | 2018-03-07 | 2021-12-17 | 汕头大学 | Eyeground image blood vessel segmentation method based on layered matting algorithm |
-
2018
- 2018-09-09 CN CN201811047463.8A patent/CN109308708B/en active Active
- 2018-09-09 CN CN202010048619.5A patent/CN111311625A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070237408A1 (en) * | 2006-04-05 | 2007-10-11 | Kabushiki Kaisha Toshiba | Image processing apparatus and image processing method |
| CN101826228A (en) * | 2010-05-14 | 2010-09-08 | 上海理工大学 | Detection method of bus passenger moving objects based on background estimation |
| CN104599502A (en) * | 2015-02-13 | 2015-05-06 | 重庆邮电大学 | Method for traffic flow statistics based on video monitoring |
| CN104748684A (en) * | 2015-04-13 | 2015-07-01 | 北方工业大学 | Visual detection method and device for crankshaft shoulder back chipping |
| CN105136070A (en) * | 2015-08-21 | 2015-12-09 | 上海植物园 | Method for utilizing electronic device to measure plant leaf area |
| CN106390285A (en) * | 2016-09-30 | 2017-02-15 | 深圳硅基仿生科技有限公司 | Charge compensation circuit, charge compensation method and artificial retina system |
Non-Patent Citations (1)
| Title |
|---|
| 王静: "视网膜假体中人工视觉信息处理及优化表达研究", 《中国博士学位论文全文数据库医药卫生科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643373A (en) * | 2019-12-31 | 2021-11-12 | 深圳硅基仿生科技有限公司 | Retina stimulator with bar vision detection |
| CN116929311A (en) * | 2023-09-19 | 2023-10-24 | 中铁第一勘察设计院集团有限公司 | Section deformation monitoring method, device and system for zoom imaging and storage medium |
| CN116929311B (en) * | 2023-09-19 | 2024-02-02 | 中铁第一勘察设计院集团有限公司 | Section deformation monitoring method, device and system for zoom imaging and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109308708B (en) | 2020-04-03 |
| CN111311625A (en) | 2020-06-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2995991B1 (en) | Contact lens pair, contact lens, and storage medium | |
| US10962808B2 (en) | Contact lens with image pickup control | |
| US20200004052A1 (en) | Apparatus and Method for Improving, Augmenting or Enhancing Vision | |
| RU2586238C2 (en) | Electronic ophthalmic lens with pupil convergence sensor | |
| CN107690617A (en) | For the technology for the vision correction for controlling wearable computing devices | |
| US10013599B2 (en) | Face detection, augmentation, spatial cueing and clutter reduction for the visually impaired | |
| CN104020582B (en) | Electronic type ophthalmic lens with eye gazing sensor | |
| CN110168562B (en) | Depth-based control method, depth-based control device and electronic device | |
| CN105072327B (en) | A kind of method and apparatus of the portrait processing of anti-eye closing | |
| CN106388770A (en) | Electronic ophthalmic lens with eye closed sensor with open eye prompt and data logging | |
| KR20170047172A (en) | Electronic ophthalmic lens with alarm clock | |
| TW201808215A (en) | Electronic ophthalmic lens with medical monitoring | |
| CN107007255A (en) | Electronic type ophthalmic lens with sleep monitor | |
| CN106774863A (en) | A kind of method that Eye-controlling focus are realized based on pupil feature | |
| CN109308708A (en) | Image processing method, device and the retina stimulator of low pixel | |
| CN110427108A (en) | Photographic method and Related product based on eyeball tracking | |
| CN106570447A (en) | Face photo sunglass automatic removing method based on gray histogram matching | |
| JP2018007792A (en) | Expression recognition diagnosis support device | |
| BR102016009093A2 (en) | EQUIPMENT FOR ACQUISITION OF 3D IMAGE DATE OF A FACE AND AUTOMATIC METHOD FOR PERSONALIZED MODELING AND MANUFACTURE OF GLASS FRAMES | |
| CN109345636B (en) | Method and device for obtaining virtual face image | |
| CN108415738A (en) | Electronic equipment, apparatus control method and Related product | |
| CN110313019A (en) | Information processing equipment, information processing method and program | |
| Barnes et al. | The role of vision processing in prosthetic vision | |
| CN111445527B (en) | Method for detecting bar-grid vision of retina stimulator | |
| CN112396667B (en) | Method for matching electrode positions of retina stimulator |
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: Chen Dawei Inventor after: Chen Zhi Inventor after: Wang Zhui Inventor after: Zhong Canwu Inventor before: Chen Dawei Inventor before: Wang Zhui Inventor before: Chen Zhi Inventor before: Zhong Canwu |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 518000 A, 4 building, 3 Ting Wei Industrial Park, 6 Baoan District Road, Xin'an, Shenzhen, Guangdong. Patentee after: Shenzhen Silicon Bionics Technology Co.,Ltd. Address before: 518000 A, 4 building, 3 Ting Wei Industrial Park, 6 Baoan District Road, Xin'an, Shenzhen, Guangdong. Patentee before: SHENZHEN SIBIONICS TECHNOLOGY Co.,Ltd. |