CN106485708A - A kind of round log method of counting based on image recognition - Google Patents
A kind of round log method of counting based on image recognition Download PDFInfo
- Publication number
- CN106485708A CN106485708A CN201610888326.1A CN201610888326A CN106485708A CN 106485708 A CN106485708 A CN 106485708A CN 201610888326 A CN201610888326 A CN 201610888326A CN 106485708 A CN106485708 A CN 106485708A
- Authority
- CN
- China
- Prior art keywords
- round log
- image
- region
- threshold
- counting based
- 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
- G06M—COUNTING MECHANISMS; COUNTING OF OBJECTS NOT OTHERWISE PROVIDED FOR
- G06M11/00—Counting of objects distributed at random, e.g. on a surface
-
- 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
- G06T2207/30108—Industrial image inspection
- G06T2207/30161—Wood; Lumber
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
A kind of method the invention discloses round log method of counting based on image recognition, in particular according to round log quantity statistics is carried out to the multi-threshold segmentation result of image.Belong to technical field of image processing.The present invention extracts background area first to pretreated picture, then using the method for multi-threshold segmentation, pretreated image is carried out with binaryzation, and extracts alternative round log region using the method finding straightway and connected region.Threshold segmentation is reused to the alternative round log region extracted, morphological operation etc. is processed, finally extract the connected region center meeting round log condition, and count.The method is not high to round log shape need, and efficiently solves situations such as outdoor shooting illumination condition is uneven, has preferable practicality and reliability.
Description
Technical field
The invention belongs to technical field of image processing is and in particular to a kind of round log method of counting based on image recognition, special
It is not the method using multi thresholds, round log region split to count round log quantity.
Background technology
Currently, during Wood Transportation, inspection to vehicle loading timber and the mode of examination are main or with people
Based on work mode, high labor intensive, inefficiency, error-prone.The automatic identification therefore counting for timber is also got over
Carry out more researchs.
Based on the Wood Identification Method of image recognition, mainly at the image by the timber more neat to arrangement
Manage and to obtain a kind of method of timber quantity.Timber counts the collection relating generally to timber image, the process of timber image, timber
Region segmentation, the step such as timber cluster.
The method counting for round log at present mainly has following several：First, the circle detection method based on Hough transform,
The method for circularity preferable border circular areas effect preferably, proposes higher requirement to the shape of log, practicality is poor.
(Chen Ke, Wu Jianping, Li Jinxiang, etc. the many circle detection method of the real-time robust [J] of one-dimension probability Hough transform. area of computer aided
Design and graphics journal, 2015 (10):1832-1841.) two, template matching method, template matching is a kind of to utilize round log mould
Plate, carries out mating the method obtaining statistical result with testing image.The amount of calculation of template matching method and amount of storage require to compare
Greatly, and the shape of template and pending object there is also difference problem, adaptability is not especially strong.(Hou Weiyan, Zhang Liwei,
Party boa, etc. a kind of bar counting measuring system based on image procossing, design and realize [J]. Chinese journal of scientific instrument, 2013,
34(5):1100-1106.) three, the method for central cluster, the method carries out center and strengthens and then center is carried out to image border
Cluster, the method is higher for the radius of timber and circularity requirement, and same adaptability is not particularly strong.4th, center is sent out
Scattered method, in the method centered on bar, successively enters line retrieval to surrounding, and progressively spreads, and until finding all bars is
Only, the method speed, but the diametric requirements for circular bar are higher it is impossible to deviation is too big.(Hou Weiyan, Hou Zhaoyang.
Lash the image recognition counting algorithm [J] of bar. instrument and meter for automation, 2016 (5) .) five, threshold segmentation method, the method passes through
Threshold segmentation is passed through to round log image, and using the operation such as corrosion, is partitioned into each round log region and is counted, but the method
Image taking is required higher, interference factor is larger.(Xue Yanbing. the counter system of bar research based on machine vision and realization
[D]. Shandong University, 2014.), in the document, using industrial camera, equalize illumination to ensure the quality of image.
Content of the invention
Goal of the invention：In order to overcome the deficiencies in the prior art, the present invention provides a kind of adaptability stronger round log
Method of counting.
Technical scheme：For achieving the above object, the technical solution used in the present invention is：
A kind of round log method of counting based on image recognition, comprises the steps of；
Step 1, carries out pretreatment to the shooting image of round log, obtains the normalized images of detection zone；
Step 2, calculates threshold value, goes out background area according to described Threshold segmentation, obtain round log region；
Step 3, dividing processing round log image being carried out including multi-threshold segmentation using several threshold values, it is partitioned into complete
Portion meets the round log region of condition；
Step 4, is screened further according to the result of step 3, obtains qualified round log region, and counts.
Further, the method for the Image semantic classification in step 1 is：Described shooting image gray processing is processed, and to
To gray level image zoom in and out, obtain the normalized images of detection zone.
Further, by described gray level image, proportional zoom to the width including length and width is 500 pixels.
Further, the threshold value determination method of the segmentation background area in step 2 is：Initially set up rectangular histogram, to straight
Side schemes the gray value that ascending statistics has reached total pixel 85%, carries out background segment with this gray value for threshold value.
Further, described step 3 is realized as follows；
Step 3-1, initially sets up with the null images of the equal size of normalized images as round log region accumulated image；
Step 3-2, chooses first threshold value and enters row threshold division to described normalized images, obtain bianry image；
Step 3-3, the round log background area image that described bianry image and step 2 are obtained carries out AND-operation, removes
The impact of background area；
Step 3-4, the image that step 3-3 is obtained carries out de-noise operation, obtains complete round log region；
Step 3-5, the image that step 3-4 is obtained, progressively scan straightway from top to bottom, find qualified straight line
Section, and the respective pixel position gray value cumulative 1 to newly-built accumulated image in step 3-1；
Step 3-6, the image zooming-out connected region again step 3-4 being obtained, extracted by constraints and meet round log
The region of condition, and in accumulated image corresponding location of pixels gray value cumulative 1；
Step 3-7, chooses several threshold values, repeated execution of steps 3-2 to step 3-6, obtains final accumulated image.
Further, in described step 3-4, the method for de-noise operation includes medium filtering, removes small area connected region
Domain.
Further, in described step 3-6, constraints includes the area of connected region, boundary rectangle length-width ratio.
Further, in described step 3-5, the length range finding straightway is 20～70 pixels；
In described step 3-6, judge that the condition that connected region adopts for round log region is：Connected region area be 500～
5000 pixels, the length-width ratio of boundary rectangle is less than 1.5, and circularity is 0.8～1；
In described step 3-7, the choosing method of several threshold values is, statistic histogram first, more ascending to rectangular histogram
Calculate the corresponding gray value that accumulated pixel number reaches 85%, then equidistantly calculate N number of threshold from 0 to this gray value
Value.
Further, described step 4 is realized as follows；
Step 4-1, chooses respective threshold, the accumulated image that step 3 obtains is split, primary segmentation goes out each round log
Region；
Step 4-2, the image that step 4-1 is obtained carries out medium filtering and operation is opened in morphology；
Step 4-3, extracts image connectivity region, according to the ginseng including connected region area, boundary rectangle length-width ratio
Number, is directly judged as round log to the region meeting condition, preserves regional center, and extracts the region of the condition of being unsatisfactory for new images
In to carry out postsearch screening；
Step 4-4, the image that step 4-3 is obtained, progressive scan meets the straight line line of length range from top to bottom again
Section, and extract line segment region；
Step 4-5, the image zooming-out connected region that step 4-4 is obtained, carry out round log conditional judgment again；To meeting bar
The region reservation region center of part；
Step 4-6, the image that step 4-4 is obtained carries out continuous corrosion, extracts connected region and justified after corroding every time
Wooden region decision, to the region reservation region center meeting round log condition；
Step 4-7, to all of regional center obtained above, distance merges less than the region of 200 pixels, exclusion
The situation that one round log is repeatedly identified, obtains final center, counts, obtains final round log number.
Further, the respective threshold chosen in described step 4-1 is 8；
Judge in step 4-3 that connected region for the condition in round log region is：Area between 50 pixels to 900 pixels, outward
The length-width ratio connecing rectangle is less than 1.5.
Beneficial effect：The round log method of counting based on image recognition that the present invention provides, has the advantages that：Reduce
Requirement to picture quality, efficiently solves outdoor pickup light according to the impact for round log segmentation, has well adapting to property
And reliability, there is higher accuracy rate.Efficiently solve the requirement to picture quality and round log shape in counting process
The problems such as shape requires.
Brief description
Fig. 1 is a kind of algorithm flow chart of the round log method of counting based on image recognition of the present invention.
Fig. 2 is to the image intercepting out and to carry out the result schematic diagram of gray processing.
Fig. 3 is the background area schematic diagram being partitioned into.
Fig. 4 (a) is the result schematic diagram of removal background area after Threshold segmentation.
Fig. 4 (b) is the image schematic diagram after segmentation result is processed using filtering, morphology etc..
Fig. 4 (c) is the result schematic diagram to segmentation image zooming-out straight line.
Fig. 5 (a) is the result schematic diagram that accumulated image is entered with row threshold division.
Fig. 5 (b) be the image after Threshold segmentation is filtered, the result schematic diagram after morphological operation.
Fig. 5 (c) is the result schematic diagram extracting straightway.
Fig. 5 (d) is recognition result schematic diagram.
Specific embodiment
Below in conjunction with the accompanying drawings the present invention is further described.
The present invention is a kind of round log method of counting based on image recognition, in particular according to the multi-threshold segmentation knot to image
The method to carry out round log quantity statistics for the fruit.The present invention extracts background area first to pretreated picture, then adopts many
The method of Threshold segmentation carries out binaryzation to pretreated image, and is carried using the method finding straightway and connected region
Take alternative round log region.Threshold segmentation is reused to the alternative round log region extracted, morphological operation etc. is processed, and finally extracts
Meet the connected region center of round log condition, and count.The method is not high to round log shape need, and efficiently solves in outdoor
Situations such as shoot illumination condition inequality, has preferable practicality and reliability.
The present invention is split to round log image according to the method for multi-threshold segmentation and statistics horizontal line section, then basis
Segmentation result is processed to image connectivity region, obtains the center of each round log, and round log number is counted, and flow process is such as
Shown in Fig. 1.
Embodiment
Step 1, carries out the normalized images that pretreatment obtains detection zone, the pretreatment of image to the shooting image of round log
Including：The intercepting of image, image gray processing, image scaling, medium filtering.
(1) image input is generally rgb format image.
(2) image interception adopts Manual interception mode, intercepts the effective coverage comprising round log.
(3) gradation conversion formula is Gray=(306*R+601*G+117*B)>>10, the such as Fig. 2 of the image after gray processing institute
Show；
(4) image scaling acquiescence is scaled using equal proportion, and the width of image is zoomed to 500.
(5) medium filtering is carried out to the image after conversion, filter the interference of noise.
Step 2, calculates appropriate threshold, is partitioned into background area, obtains round log region；Calculate threshold value method be：
Count the rectangular histogram of round log gray level image first, according to statistic histogram, calculate accumulated pixel number according to gray value is ascending
Amount reaches the corresponding gray value of whole pixel quantities more than 85%, and enters row threshold division with this gray value to image, extracts
Background area, extracts background area result as shown in Figure 3；
Step 3, carries out multi-threshold segmentation etc. using multiple appropriate threshold and processes, be partitioned into and all meet bar to round log image
The round log region of part.
Step 3-1, initially sets up null images sumImg (whole grey scale pixel values are 0) with the equal size of normalized images
As round log region accumulated image, the region later splitting all is accumulated in above this image, every time add up 1；
Step 3-2, chooses first appropriate threshold and enters row threshold division to image, obtain bianry image.First threshold value
Computational methods adopt equisection method to calculate, point N section such as threshold value of background segment that step 2 will be calculated, N divides for multi thresholds
The number of times cutting, you can try to achieve each threshold value, N of the present invention takes 30；
Step 3-3, the round log background area image that the bianry image that step 3-2 is partitioned into is obtained with step 2 is carried out
AND-operation, removes the impact of background area, shown in result such as Fig. 4 (a)；
Step 3-4, the image that step 3-3 is obtained carries out medium filtering, removes noise jamming, and removes small area connection
Region operates, and the connected region threshold value that the present invention chooses is 500 pixels, the connected region less than this threshold value is removed, obtains relatively
For complete round log region, shown in result such as Fig. 4 (b)；
Step 3-5, the image that step 3-4 is obtained, progressively scan straightway from top to bottom, find length and arrive in 20 pixels
Line segment between 70 pixels, and in the respective pixel position of newly-built accumulated image sumImg of step 3-1 cumulative 1, extract line segment
Shown in result such as Fig. 4 (c)；
Step 3-6, the image zooming-out connected region again step 3-4 being obtained, by the area of connected region, external square
The constraintss such as shape, extract the region meeting round log condition, and the corresponding pixel of accumulated image sumImg built in step 3-1
Position cumulative 1.Constraints is：
(1) area is between 500 pixels to 5000 pixels.
(2) the long side of boundary rectangle and the ratio of minor face are less than 1.5.
(3) circularity (4*Pi*Area/C2, wherein Area is area, and C is girth) and between 0.8 to 1.
Step 3-7, according to step 3-2 selected threshold method, chooses next threshold value, repeated execution of steps 3-2 is to step
3-6, obtains final accumulated image.
Step 4, is screened further according to the result of step 3, obtains qualified round log region, and counts.
Step 4-1, selected threshold 8 is split to accumulated image sumImg that step 3 obtains, and primary segmentation goes out round log area
Domain, shown in result such as Fig. 5 (a).
Step 4-2, the image that the template of selection 5*5 obtains to step 4-1 carries out medium filtering, and the circle being 3 with radius
Shape template carries out 2 morphology and opens operation to image, the round log that breaking part is sticked together, shown in result such as Fig. 5 (b).
Step 4-3, extracts image connectivity region, calculates connected region area, and boundary rectangle, to meet the constraint condition
Region be directly judged as round log, preserve regional center, the extracted region being unsatisfactory for constraints be unsatisfactory for the region of condition
To in new images to carry out postsearch screening, constraints is as follows：
(1) connected region area is between 50 pixels to 900 pixels.
(2) length-width ratio of boundary rectangle is less than 1.5.
Step 4-4, the area image of the postsearch screening that step 4-3 is obtained, progressive scan length exists from top to bottom again
Straight-line segment between 30 to 70, and extract qualified line segment, shown in result such as Fig. 5 (c).
Step 4-5, the image zooming-out connected region that step 4-4 is obtained, carried out using the constraints of step 4-3 again
Screening, to qualified region reservation region center.
Step 4-6, takes the image that the circular shuttering that radius is 3 obtains to step 4-4 to carry out continuous 5 corrosion, rotten every time
Extract region after erosion and carry out round log region decision, to the round log region reservation region center meeting step 4-3 condition.
Step 4-7, to all of regional center obtained above, calculates distance between centers, distance is entered less than 25 pixels
Row center merges, the situation that one round log of exclusion is repeatedly identified, obtains final center, shown in result such as Fig. 5 (d).
Step 4-8, counts, obtains final round log number.
In sum, the present invention passes through multi-threshold segmentation, statistics horizontal line section and the method calculating connected region attribute
Achieve the round log count protocol based on image：
(1) method utilizing Threshold segmentation, simply efficiently, decreases the dependence condition for round log shape, has more preferably
The suitability.
(2) method choosing multi-threshold segmentation, can be with effectively solving because the uneven illumination producing when shooting be for segmentation
When the impact that produces.
(3) using the method being divided using straight line, can be effectively separate by Threshold segmentation posterior synechiae round log together,
Count beneficial to round log below.
(4) in counting screening process, using postsearch screening, preferably ensure that the accuracy of counting, and for
The merging verification of termination fruit, preferably prevents from examining because the error of segmentation produces, missing inspection more, increased the reliability of system.
The present invention can reach more than 90% for the accuracy rate that round log counts, and single gate time is less than 3s, for circle
The adaptability of wooden shape there has also been significant raising, can go out similar round round log with effective detection, and decrease shooting condition
Impact, has higher practicality.
The above be only the preferred embodiment of the present invention it should be pointed out that：Ordinary skill people for the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (10)
1. a kind of round log method of counting based on image recognition it is characterised in that：Comprise the steps of；
Step 1, carries out pretreatment to the shooting image of round log, obtains the normalized images of detection zone；
Step 2, calculates threshold value, goes out background area according to described Threshold segmentation, obtain round log region；
Step 3, dividing processing round log image being carried out including multi-threshold segmentation using several threshold values, it is partitioned into all full
The round log region of sufficient condition；
Step 4, is screened further according to the result of step 3, obtains qualified round log region, and counts.
2. the round log method of counting based on image recognition according to claim 1 it is characterised in that：Image in step 1
The method of pretreatment is：Described shooting image gray processing is processed, and the gray level image obtaining is zoomed in and out, detected
The normalized images in region.
3. the round log method of counting based on image recognition according to claim 2 it is characterised in that：By described gray processing figure
As the proportional zoom including length and width to width is 500 pixels.
4. the round log method of counting based on image recognition according to claim 1 it is characterised in that：Segmentation in step 2
The threshold value determination method of background area is：Initially set up rectangular histogram, ascending statistics has reached total pixel to rectangular histogram
85% gray value, carries out background segment with this gray value for threshold value.
5. the round log method of counting based on image recognition according to claim 1 it is characterised in that：Described step 3 according to
Following method is realized；
Step 3-1, initially sets up with the null images of the equal size of normalized images as round log region accumulated image；
Step 3-2, chooses first threshold value and enters row threshold division to described normalized images, obtain bianry image；
Step 3-3, the round log background area image that described bianry image and step 2 are obtained carries out AND-operation, removes background
The impact in region；
Step 3-4, the image that step 3-3 is obtained carries out de-noise operation, obtains complete round log region；
Step 3-5, the image that step 3-4 is obtained, progressively scan straightway from top to bottom, find qualified straightway,
And the respective pixel position gray value cumulative 1 to newly-built accumulated image in step 3-1；
Step 3-6, the image zooming-out connected region again step 3-4 being obtained, extracted by constraints and meet round log condition
Region, and accumulated image corresponding location of pixels gray value add up 1；
Step 3-7, chooses several threshold values, repeated execution of steps 3-2 to step 3-6, obtains final accumulated image.
6. the round log method of counting based on image recognition according to claim 5 it is characterised in that：In described step 3-4,
The method of de-noise operation includes medium filtering, removes small area connected region.
7. the round log method of counting based on image recognition according to claim 5 it is characterised in that：In described step 3-6,
Constraints includes the area of connected region, boundary rectangle length-width ratio.
8. the round log method of counting based on image recognition according to claim 5 it is characterised in that：In described step 3-5,
The length range finding straightway is 20～70 pixels；
In described step 3-6, judge that the condition that connected region adopts for round log region is：Connected region area is 500～5000
Pixel, the length-width ratio of boundary rectangle is less than 1.5, and circularity is 0.8～1；
In described step 3-7, the choosing method of several threshold values is, statistic histogram first, the more ascending calculating to rectangular histogram
Go out the corresponding gray value that accumulated pixel number reaches 85%, then equidistantly calculate N number of threshold value from 0 to this gray value.
9. the round log method of counting based on image recognition according to claim 1 it is characterised in that：Described step 4 according to
Following method is realized；
Step 4-1, chooses respective threshold, the accumulated image that step 3 obtains is split, primary segmentation goes out each round log area
Domain；
Step 4-2, the image that step 4-1 is obtained carries out medium filtering and operation is opened in morphology；
Step 4-3, extracts image connectivity region, according to the parameter including connected region area, boundary rectangle length-width ratio, right
The region meeting condition is directly judged as round log, preserves regional center, and extract the region of the condition of being unsatisfactory in new images with
Carry out postsearch screening；
Step 4-4, the image that step 4-3 is obtained, progressive scan meets the straight-line segment of length range from top to bottom again, and
Extract line segment region；
Step 4-5, the image zooming-out connected region that step 4-4 is obtained, carry out round log conditional judgment again；To qualified
Region reservation region center；
Step 4-6, the image that step 4-4 is obtained carries out continuous corrosion, extracts connected region and carry out round log area after corroding every time
Domain judges, to the region reservation region center meeting round log condition；
Step 4-7, to all of regional center obtained above, distance merges less than the region of 200 pixels, excludes one
The situation that round log is repeatedly identified, obtains final center, counts, obtains final round log number.
10. the round log method of counting based on image recognition according to claim 9 it is characterised in that：In described step 4-1
The respective threshold chosen is 8；
Judge in step 4-3 that connected region for the condition in round log region is：Area between 50 pixels to 900 pixels, external square
The length-width ratio of shape is less than 1.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610888326.1A CN106485708B (en) | 2016-10-11 | 2016-10-11 | A kind of round log method of counting based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610888326.1A CN106485708B (en) | 2016-10-11 | 2016-10-11 | A kind of round log method of counting based on image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106485708A true CN106485708A (en) | 2017-03-08 |
| CN106485708B CN106485708B (en) | 2018-12-28 |
Family
ID=58269899
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610888326.1A Active CN106485708B (en) | 2016-10-11 | 2016-10-11 | A kind of round log method of counting based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106485708B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108010076A (en) * | 2017-11-23 | 2018-05-08 | 南通大学 | A kind of end face appearance modeling method towards intensive industry bar image detection |
| CN108629774A (en) * | 2018-05-11 | 2018-10-09 | 电子科技大学 | A kind of annular object method of counting based on hough-circle transform |
| CN109087276A (en) * | 2018-05-17 | 2018-12-25 | 苏州斯玛维科技有限公司 | The component of radioscopic image based on SMT charging tray counts automatically and localization method |
| CN109741303A (en) * | 2018-12-25 | 2019-05-10 | 浙江中烟工业有限责任公司 | A kind of filter stick counting method and apparatus based on Image Acquisition and identification |
| CN109801260A (en) * | 2018-12-20 | 2019-05-24 | 北京海益同展信息科技有限公司 | The recognition methods of livestock number and device |
| CN109977960A (en) * | 2019-04-03 | 2019-07-05 | 杭州深数科技有限公司 | Rollway information acquisition method neural network based, system, device |
| US10586347B1 (en) | 2018-09-17 | 2020-03-10 | Datalog, LLC | Log scaling system and related methods |
| CN111260723A (en) * | 2020-04-30 | 2020-06-09 | 长沙海贝智能科技有限公司 | Barycenter positioning method of bar and terminal equipment |
| CN113160181A (en) * | 2021-04-25 | 2021-07-23 | 湖南九九智能环保股份有限公司 | Stacking profile counting method based on image recognition |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006178694A (en) * | 2004-12-22 | 2006-07-06 | Hitachi Software Eng Co Ltd | Method for detecting candidate of constitutional segment of road mark image, and program capable of detecting candidate of constitutional segment of road mark image |
| CN103810522A (en) * | 2014-01-08 | 2014-05-21 | 中国农业大学 | Counting method and device for corn ear grains |
| CN104240204A (en) * | 2014-09-11 | 2014-12-24 | 镇江苏仪德科技有限公司 | Solar silicon wafer and battery piece counting method based on image processing |
| CN104778711A (en) * | 2015-04-27 | 2015-07-15 | 厦门大学 | Cell counting and positioning method at cleavage stage |
| CN105718989A (en) * | 2014-11-30 | 2016-06-29 | 中国科学院沈阳自动化研究所 | Bar counting method based on machine vision |
-
2016
- 2016-10-11 CN CN201610888326.1A patent/CN106485708B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2006178694A (en) * | 2004-12-22 | 2006-07-06 | Hitachi Software Eng Co Ltd | Method for detecting candidate of constitutional segment of road mark image, and program capable of detecting candidate of constitutional segment of road mark image |
| CN103810522A (en) * | 2014-01-08 | 2014-05-21 | 中国农业大学 | Counting method and device for corn ear grains |
| CN104240204A (en) * | 2014-09-11 | 2014-12-24 | 镇江苏仪德科技有限公司 | Solar silicon wafer and battery piece counting method based on image processing |
| CN105718989A (en) * | 2014-11-30 | 2016-06-29 | 中国科学院沈阳自动化研究所 | Bar counting method based on machine vision |
| CN104778711A (en) * | 2015-04-27 | 2015-07-15 | 厦门大学 | Cell counting and positioning method at cleavage stage |
Non-Patent Citations (1)
| Title |
|---|
| 邓凯: "针对钢筋计数系统的图像处理方法研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108010076A (en) * | 2017-11-23 | 2018-05-08 | 南通大学 | A kind of end face appearance modeling method towards intensive industry bar image detection |
| CN108010076B (en) * | 2017-11-23 | 2020-03-31 | 南通大学 | End face appearance modeling method for intensive industrial bar image detection |
| CN108629774A (en) * | 2018-05-11 | 2018-10-09 | 电子科技大学 | A kind of annular object method of counting based on hough-circle transform |
| CN109087276A (en) * | 2018-05-17 | 2018-12-25 | 苏州斯玛维科技有限公司 | The component of radioscopic image based on SMT charging tray counts automatically and localization method |
| CN109087276B (en) * | 2018-05-17 | 2022-02-01 | 苏州斯玛维科技有限公司 | Automatic counting and positioning method for components based on X-ray image of SMT tray |
| US10825192B2 (en) | 2018-09-17 | 2020-11-03 | Datalog, LLC | Log scaling system and related methods |
| US12008498B2 (en) | 2018-09-17 | 2024-06-11 | Datalog, LLC | Log scaling system and related methods |
| US11694138B2 (en) | 2018-09-17 | 2023-07-04 | Datalog, LLC | Log scaling system and related methods |
| US10586347B1 (en) | 2018-09-17 | 2020-03-10 | Datalog, LLC | Log scaling system and related methods |
| US11393120B2 (en) | 2018-09-17 | 2022-07-19 | Datalog, LLC | Log scaling system and related methods |
| CN109801260A (en) * | 2018-12-20 | 2019-05-24 | 北京海益同展信息科技有限公司 | The recognition methods of livestock number and device |
| CN109741303A (en) * | 2018-12-25 | 2019-05-10 | 浙江中烟工业有限责任公司 | A kind of filter stick counting method and apparatus based on Image Acquisition and identification |
| CN109977960A (en) * | 2019-04-03 | 2019-07-05 | 杭州深数科技有限公司 | Rollway information acquisition method neural network based, system, device |
| CN111260723B (en) * | 2020-04-30 | 2020-08-14 | 长沙海贝智能科技有限公司 | Barycenter positioning method of bar and terminal equipment |
| CN111260723A (en) * | 2020-04-30 | 2020-06-09 | 长沙海贝智能科技有限公司 | Barycenter positioning method of bar and terminal equipment |
| CN113160181A (en) * | 2021-04-25 | 2021-07-23 | 湖南九九智能环保股份有限公司 | Stacking profile counting method based on image recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106485708B (en) | 2018-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106485708A (en) | A kind of round log method of counting based on image recognition | |
| CN111382704B (en) | Vehicle line pressing violation judging method and device based on deep learning and storage medium | |
| CN101515325B (en) | Character extracting method in digital video based on character segmentation and color cluster | |
| CN109636824B (en) | Multi-target counting method based on image recognition technology | |
| CN114998350A (en) | Stone defect detection method based on image processing | |
| CN104835179B (en) | Based on the adaptive improvement ViBe background modeling methods of dynamic background | |
| CN104537651B (en) | Proportion detecting method and system for cracks in road surface image | |
| CN107610104A (en) | Crack detecting method at a kind of elevator compensation chain R based on machine vision | |
| CN106548160A (en) | A kind of face smile detection method | |
| CN105139386A (en) | Image processing method for quickly and automatically detecting soldered dot unsatisfied products of electric connectors | |
| CN101847265A (en) | Method for extracting moving objects and partitioning multiple objects used in bus passenger flow statistical system | |
| CN105005813B (en) | Method of counting and system are analyzed in insect pest | |
| CN109087363B (en) | HSV color space-based sewage discharge detection method | |
| WO2017166597A1 (en) | Cartoon video recognition method and apparatus, and electronic device | |
| WO2021135675A1 (en) | Automatic bar counting method and apparatus | |
| CN105335741A (en) | Smudged serial number classification method and system | |
| CN105719275A (en) | Parallel combination image defect segmentation method | |
| CN103778627A (en) | Sea oil spill detection method based on SAR image | |
| CN104700420A (en) | Ellipse detection method and system based on Hough conversion and ovum identification method | |
| CN108280838A (en) | A kind of intermediate plate tooth form defect inspection method based on edge detection | |
| CN112560941A (en) | Coal and gangue identification method based on image detection | |
| CN111476804A (en) | Method, device and equipment for efficiently segmenting carrier roller image and storage medium | |
| CN107798304A (en) | A kind of method of fast video examination ＆ verification | |
| CN105701490B (en) | A kind of container number adaptive location method based on image entropy | |
| CN115797361A (en) | Aluminum template surface defect detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |