CN1207683C - Ordered picture splicing method in seed grain-space detection - Google Patents

Abstract

The present invention relates to a method for detecting grain spaces of seeds, particularly to a method for splitting sequential images in the detection of grain spaces of seeds. The method for splitting sequential images in the detection of grain spaces of seeds is characterized in that the method comprises the following steps that (1), a plurality of marks with different sizes and the same interval are made on a recording adhesive tape; (2), a first complete mark of the overlapping region of two adjacent images of the sequential images acquired by a computer is used as a reference so as to find the form center of the first complete mark; (3), pixels of the former image are taken up to the form center of the first complete mark of the image overlapping region; (4), pixels of the latter image are taken from the form center of the same mark of the image overlapping region; (5), in the two adjacent images, grain spaces of adjacent seeds which are not in the overlapping region are the sum of horizontal distance between seeds in the first image and the center form of the image mark, and horizontal distance between the form center of the image mark in the second image and seeds in the image.

Description

Seed grain sequence image joining method in detect

Technical field

The present invention relates to the joining method of a kind of seed grain distance detecting method, particularly seed grain sequence image in detect.

Background technology

Based on the detection technique of Flame Image Process and analysis is that collection computer technology, image technique, automatic control technology are the method for comprehensive detection of one, have directly, fast, true, reliable characteristics, be a kind of contactless direct detecting method.Along with the specialization of image processing techniques and the raising of computing power price ratio, this technology is widely used in the agricultural engineering field.

The detection of seed grain distance is the committed step of carrying out the precision drill Performance Detection, present seed grain both domestic and external is apart from detecting mainly based on the Photoelectric Detection means, calculates the spacing of seed by writing down speed that time interval that two seeds fall and land wheel advance.Its method belongs to the actual pitch that indirect detection can not directly reflect seed.

Summary of the invention

The object of the present invention is to provide a kind of seed grain distance detecting method based on Flame Image Process and analytical technology, this method can realize that three planters test simultaneously, shortens test period; The data of measuring are handled immediately, provided the performance index such as qualified index, replay index, broadcast leakage index of precision drill, draw grain automatically apart from distribution histogram.

Seed grain sequence image joining method in detect, its characteristics are:

This method comprises the steps:

(1) on the record adhesive tape, makes the alternate mark of several equidistant sizes;

(2) first complete mark of the overlapping region of two frame adjacent images of the sequence image of being gathered with computing machine is as the criterion and finds its centre of form;

(3) centre of form of getting first complete mark in this doubling of the image zone of the pixel in the former frame image is ended;

(4) pixel in one two field picture of back begins to get from the centre of form of the same tag in this doubling of the image zone;

In (5) the two frame adjacent images not at the grain of the adjacent seed of overlapping region apart from the horizontal range sum that is seed in the centre of form of the seed image tagged in the horizontal range of the centre of form of this image tagged and second two field picture in first two field picture and this image.

This method is based on the seed grain distance detecting method of Flame Image Process and analytical technology, and precision drill seeding situation dynamic detection technology when having studied the detection testing table and moving with friction speed has realized that three planters test simultaneously, has shortened test period greatly.By processing and analysis to the sequence image gathered, utilize image matching technology based on mark, realized the measurement of a plurality of seed spacings.And the data of measuring are handled immediately, provide the performance index such as qualified index, replay index, broadcast leakage index of precision drill, draw grain automatically apart from distribution histogram.

Description of drawings

Fig. 1 is a test system architecture block diagram of the present invention

Fig. 2 is the connection diagram of camera of the present invention and image card

Fig. 3 is an acquired original image of the present invention

Fig. 4 is image behind the thresholding of the present invention

Fig. 5 is an image mosaic principle of the present invention

Fig. 6 is that grain of the present invention is apart from distribution histogram

Embodiment

Please refer to Fig. 1, seed grain apart from detecting principle is: the static rubber strip top that is installed in of seeder or planter (feed mechanism for seed), drag down at buncher, and the relative seeder of rubber strip moves horizontally.During test, seed in the kind case drops through feed mechanism for seed, discharging tube and (is scribbling lubricating oil on the adhesive tape on the adhesive tape, so that seed is clung, make itself and adhesive tape that relative displacement not take place), adhesive tape is gathered the image in certain breadth in real time by CCD measuring system camera the time, cut apart, change through image card, make computing machine analysis, calculate, handle, thereby enforcement is to the measurement of the seed grain number in seed grain distance or the unit distance.The inhomogeneity evaluation result of given planting with sowing machine.System program implements to coordinate control to line speed and image range finding speed simultaneously.

The composition of this detection system module:

1, steering order module: major equipment sequence switch control, rubber strip translational speed, the control of feed mechanism for seed rotary speed instruction.

2, image capture module: according to specified tape speed, the size and the measuring accuracy of balance overlapping region are with the image buffer storage district of sequence image collecting computer.

3, image is in advance whole cuts apart module: to Flame Image Process, with seed and background separately, and suffer seed and discern squeezing.

4, data analysis, computing module: the number of seed in the spacing of calculating seed or the certain-length interval.Require to analyze various performance parameters according to seed grain apart from detecting.

5, post-processing module: deposit test findings in database, generate test findings form and printing as requested.

6, system's automatic control module: the instruction that the receiving center computing machine sends, finish the each several part of total system and coordinate control.

System hardware structure:

Measuring system will be carried out the test of three planters simultaneously, finish the detection of triplex row seed grain distance, therefore selected a coloured image card of supporting that RGB three-component vision signal is independently gathered for use, with of 3 standard R, Gs, the B component input of 3 synchronous black and white independent video sources, finish the storage of data according to 8 * 3=24 position mode as image card.

Color Image Acquisition card: OK_RGB10;

CCD black and white industrial camera: WAT-505EX;

Video distributor was one in one minute four

Camera lens: SE1614 (16mm, 1/10000s);

Computing machine: Benyue 2000 PIII 933 of association.

Fig. 2 is the connection diagram of camera of the present invention and image card.

Image acquisition and analysis:

Image acquisition:

This detection system is a dynamic measuring system.During test, tape speed is respectively 0.5m/s, 1m/s, 1.5m/s, 2m/s, 2.5m/s, 3m/s.Overlapping the least possible in order to make before and after the sequence image between two frames, do not influence the precision of splicing again, solving dynamic image acquisition is the problem that must solve.

It is with the mode that scans line by line in sequence that video camera obtains image formation vision signal.Two of staggered scanning branch odd evens are carried out, and strange sweep trace is inserted in the middle of the idol adjacent scanning lines uniformly.Odd even two occasions are a frame.When the adhesive tape linear velocity was up to 3m/s, general video camera 40ms adopted a frame, and target will move 3 * 40=120mm during this, and seed becomes the rectangular of a distortion in the image of gathering, and will badly influence later Flame Image Process and analysis.Native system adopts the camera of charged sub-shutter, gets rid of image blurring owing to what move and brought by shortening the time shutter.

The image of camera collection divides parity field, and the time interval between strange, idol is 20ms, and in the two field picture of being made up of the odd even two field picture, same objective body but has two pictures, and which specifically adopts as seed, can't differentiate.The OK_RGB10 image pick-up card can be gathered single game, single frames, several frames in interval, successive frame etc., accurately show up, utilize this characteristic of image card, adopt by an acquisition order and deposit image, be unit with the field when handling, so just can guarantee that a seed has only a picture in a two field picture.

Because the speed of camera collection image is 25 frames/s, the prestissimo that adhesive tape moves is 3m/s, takes in the time of a two field picture, and tape shifter is crossed 120mm.Take if press the intrinsic frequency of camera, between two two field pictures of front and back a large amount of redundant informations will be arranged.The frame period that utilizes the OK-RGB10 image pick-up card to provide is provided with function okSetCaptureParam (hBoard, CAPTURE_INTERVAL, INTERVAL), according to different speed different frame periods is set, reduce data volume reaching, the purpose of the correct splicing of two two field pictures before and after guaranteeing simultaneously to realize.The computing formula of frame period is as follows:

$\mathrm{INTERVAL}=\frac{200}{V\×20}-1,$ INTERVAL---frame period wherein, V---tape speed

Seed separates:

Native system has adopted the image threshold split plot design to realize separating of background and seed.The gray-scale value of pixel is a seed greater than threshold value, and less than threshold value is background.Based on the image of gathering fixing light illumination is arranged, gather environment and also do not change, therefore adopted fixing threshold value, according to result of experiment repeatedly, threshold value is made as 50 and has reached good effect.As the original image of Fig. 3 for gathering, Fig. 4 is the image behind two thresholdings.

Among Fig. 3, Fig. 4, black part is divided into background, and the brightest point is a green channel, and the darkest point is a blue channel, and what brightness was placed in the middle is red channel.Size alternate point in image top is a mark, and mark is positioned at green channel.

Image kind behind the thresholding as can be seen, background and seed are separated fully, size alternate point in image top is the image mosaic mark, the big zone of three kinds of colors of middle part RGB is respectively the seed that three planters broadcast, and adopting for convenience here, hand dipping adopts the scraps of paper to replace seed.

In order to discern different seed demands mark is carried out in the zone that is communicated with among the last figure.Give the purpose that different marks has just reached seed identification with different zones.Here adopted the method for element marking: image is from left to right carried out four from top to bottom be communicated with scanning (scanning area does not comprise the mark zone),, just move on to next scanning position if the gray-scale value of current pixel is 0.If the gray-scale value of current pixel is 255, check two neighbour's pixels (according to the scanning sequence that is adopted, these two neighbour's pixels are processed when scanning arrives current pixel) of its left side and top.Need to consider four kinds of situations: (1) their gray-scale value all is 0, gives new mark of current pixel; (2) having only a gray-scale value is 255, just the mark of this pixel is composed to current pixel; (3) their gray-scale value all is 255 and has identical mark, just this mark is composed to current pixel; (4) their gray-scale value all is 255 and has different marks, just one of them mark is composed to current pixel, and is marked and show this two mark equivalences ^{[2] [3]}Rescan image at last, each mark is replaced with the of equal value right mark in its place.Three passages of RGB scan respectively, and the seed of each passage is done for oneself one group, and are irrelevant with other passages.

Plant determining of subcenter:

After the mark that has carried out seed, the pixel that mark is the same is just thought and is belonged to same seed.For different zones, as long as obtain its centre of form, just this center as seed.We only are concerned about that two seeds are broadcasting the projector distance of capable center line, so as long as obtain the coordinate of this direction.As true origin, is to the right the X-axis forward with the upper left corner of image, is downwards the Y-axis forward, and the X coordinate of every kind subcenter is institute and asks, and its computing formula is as follows ^[4]:

$\stackrel{\‾}{x}=\frac{\underset{t=0}{\overset{n-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{m-1}{\mathrm{\Σ}}}\mathrm{jB}[i,j]}{A}$ Wherein $A=\underset{i=0}{\overset{n-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{m-1}{\mathrm{\Σ}}}B[i,j],$ B[I, j] be the mark of this point

The note value

Carry out the X coordinate figure that above-mentioned computing just obtains the centre of form of all seeds respectively for different.

Sequence image splicing principle:

Because being dynamic real-time, this detection system detects, the stepless change in the 0.5m/s---3m/s scope of its line speed, detecting testing table is that 16m is long, 1.97m it is wide, finish the real-time detection of three seeders (planter) seeding accuracy simultaneously, 250 of each planting with sowing machine, 30mm---150mm does not wait its intergranular apart from being respectively, the length that its travelling belt moves is 7500mm-----37500mm, if the capture scope with a two field picture is that 270mm length * 250mm is wide, the images acquired number is 30 width of cloth---150 width of cloth, overlapping the least possible in order to make before and after the sequence image between two frames, do not influence the precision of splicing again, solving dynamic image acquisition and sequence image splicing problem becomes key.

The key issue that sequence image detects is exactly the splicing between the image, and accurately whether images match the precision important influence to measuring.Native system has adopted on adhesive tape and has marked, and with the method for this mark as the standard of front and back image two two field pictures couplings.Do like this and to be based on following principle:

Be the two frame adjacent images that computing machine is gathered shown in Fig. 5 (a), realize the measurement of two the seed spacings in border, promptly as Fig. 5 (b) two two field pictures are stitched together, but calculated amount is quite big like this, because will move a large amount of pixels, algorithm is also more numerous and diverse.In two frame figure, be as the criterion respectively as Fig. 5 (c) and find the center of mark with mark, pixel is got this just end in the first frame figure, center from mark in the second frame figure begins the capture element, be equivalent to two two field pictures are stitched together, remove the work of moving pixel in a large number that image mosaic will be done from, and reached the effect of image mosaic.For asking L, can measure L1 and L2 respectively, obtain L=L1+L2.For drill for sowing in lines, get certain distance equally and measure the interior seed grain number of this section, for the intersection on border, as long as two two field pictures are as the criterion with same mark.

The frame period that doubling of the image zone is provided with when gathering and deciding, its computing formula is as follows:

The overlapping region of image and former frame image: [0, (270-(INTERVAL+1) * 20 * V) * 768 ÷ 270]

The overlapping region of image and back one two field picture: [768-(270-(INTERVAL+1) * 20 * V) * 768 ÷ 270,768]

The program implementation algorithm:

A, find first the complete mark of overlapping region in the previous frame image

B, find first the complete mark of overlapping region in the next frame image

C, judge that two mark size are whether consistent

If the d unanimity is just thought indicia matched

If e is inconsistent, second mark just getting previous frame or next frame mates, till matching.Considered the influence of tape speed in the algorithm of above-mentioned indicia matched, be no more than a mark spacing as long as tape speed changes the variable in distance that adhesive tape moves in 20ms that causes, above-mentioned algorithm all is effective.

Measurement result:

Table 1 has intercepted the one group measurement of a part when tape speed is 0.5m/s and has used measured result and the hand dipping value of native system.We use this system that same group of seed carried out twice experiment, so that the consistance as a result of using native system to detect is compared.

The qualified index of this experiment is 72.0, and the replay index is 17.89, and the broadcast leakage index is 10.53, and mean value is 0.98, and standard deviation is 0.23.As can be seen from Table 1, the measurement result of native system and hand dipping result have good consistance, and deviation meets the seeder performance test and detects requirement within ± 2mm.And further find twice contrast experiment that the use native system carries out, result's consistance is also better.The histogram of painting 6 in full accord with the actual distribution of measurement result and seed.

Hand dipping result (mm)	Native system measurement result 1 (mm)	Deviation (mm)	Native system measurement result 2 (mm)	Deviation (mm)
					26.90	27.07	-0.17	27.07	-0.17
91.00	90.352	0.65	90.352	0.65
					120.90	120.586	0.31	120.938	-0.04
7.50	6.68	0.82	7.031	0.47
					124.80	124.102	0.70	124.102	0.70
115.50	115.313	0.19	115.313	0.19
					73.00	73.477	-0.48	73.125	-0.13
54.80	53.789	1.01	54.141	0.66
					45.50	45.352	0.15	45.352	0.15
90.50	90.352	0.15	90	0.50

Claims (1)

1, a kind of seed grain sequence image joining method in detect is characterized in that: this method comprises the steps:

(1) on the record adhesive tape, makes the alternate mark of several equidistant sizes, the static adhesive tape top that is installed in of seeder or planter, drag down at buncher, the relative seeder of adhesive tape moves horizontally, and the seed of planting in the case drops on adhesive tape through feed mechanism for seed, discharging tube, scribble lubricating oil on the adhesive tape, so that seed is clung, make itself and adhesive tape that relative displacement not take place, adhesive tape is by the measuring system camera time, camera and image card cooperate, in real time images acquired;

(2) adopt the image threshold split plot design to realize separating of background and seed, the gray-scale value of pixel is a seed greater than threshold value, and less than threshold value is background;

(3) adopt the method for element marking that mark is carried out in the zone that is communicated with among the figure, give different marks to discern different seeds different zones; Image is from left to right carried out four from top to bottom be communicated with scanning, scanning area does not comprise the mark zone, is 0 as the gray-scale value of current pixel, just moves on to next scanning position; Gray-scale value as current pixel is 255, checks two neighbour's pixels of its left side and top, comprises that four kinds of their gray-scale values of situation: A. all are 0, gives new mark of current pixel; B. having only a gray-scale value is 255, just the mark of this pixel is composed to current pixel; C. their gray-scale value all is 255 and has identical mark, just this mark is composed to current pixel; D. their gray-scale value all is 255 and has different marks, just one of them mark is composed to current pixel, and marking shows this two mark equivalences; Rescan image at last, each mark is replaced with the of equal value right mark in its place;

(4) pixel that mark is the same belongs to same grain seed, obtaining the center of its centre of form as seed for different zones, as true origin, is to the right the X-axis forward with the image upper left corner, be the Y-axis forward downwards, the X coordinate Calculation formula that calculates each kind subcenter is as follows:

$\stackrel{\‾}{x}=\frac{\underset{i=0}{\overset{n-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{m-1}{\mathrm{\Σ}}}\mathrm{jB}[i,j]}{A}$

Wherein $A=\underset{i=0}{\overset{n-1}{\mathrm{\Σ}}}\underset{j=0}{\overset{m-1}{\mathrm{\Σ}}}B[i,j],$ B[i, j] be the mark value of this point;

(5) centre of form of getting first complete mark in this doubling of the image zone of the pixel in the former frame image is ended;

(6) pixel in one two field picture of back begins to get from the centre of form of the same tag in this doubling of the image zone;

In (7) the two frame adjacent images not at the grain of the adjacent seed of overlapping region apart from the horizontal range sum that is seed in the centre of form of the seed image tagged in the horizontal range of the centre of form of this image tagged and second two field picture in first two field picture and this image;

The frame period that described doubling of the image zone is provided with when gathering and deciding, its computing formula is as follows:

The overlapping region of image and former frame image:

[0，(270-(INTERVAL+1)*20*V)*768/270]

The overlapping region of image and back one two field picture:

[768-(270-(INTERVAL+1)*20*V)*768/270，768]

Wherein: INTERVAL is a frame period, and V is the adhesive tape travelling speed

The program implementation algorithm:

A finds first complete mark of overlapping region in the previous frame image;

B finds first complete mark of overlapping region in the next frame image;

C judges whether two mark size are consistent;

If the d unanimity is just thought indicia matched;

If e is inconsistent, second mark just getting previous frame or next frame mates, till matching.

Images