CN1434959A - Image processing device and elevator mounting it thereon - Google Patents
Image processing device and elevator mounting it thereon Download PDFInfo
- Publication number
- CN1434959A CN1434959A CN00819023.2A CN00819023A CN1434959A CN 1434959 A CN1434959 A CN 1434959A CN 00819023 A CN00819023 A CN 00819023A CN 1434959 A CN1434959 A CN 1434959A
- Authority
- CN
- China
- Prior art keywords
- image
- wheelchair
- zone
- mentioned
- face
- 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
Images
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
An image processing device having a function of detecting wheel-chair riders; which comprises a TV camera installed so as to monitor a hall portion, where riders are waiting, from the vicinity of the elevator door, a face area extracting means, a wheel-chair area estimating means, and a wheel-chair presence/absence judging means; and which operates such that the face area extracting means specifies the face areas of one or a plurality of riders based on an image picked up by the TV camera and detects their positions on the image, the wheel-chair area estimating means estimates, from the face-area positions, areas where wheel chairs are assumed to exist on the image if riders are on wheel chairs, and the wheel chair presence/absence judging means checks the insides of the estimated wheel-chair presence areas and judges the presence/absence of wheel chairs.
Description
Technical field
The present invention relates to judge the passenger's of elevator etc. the image-processing system of state, be mounted with the elevator of this image-processing system.
Background technology
In service at elevator is importantly distinguished and takes advantage of the passenger of seat in cab to take advantage of the handicapped person of seat on wheelchair, or normal health person.Its reason is: when common healthy person uses elevator, be common operational mode, the occasion in that wheelchair occupant is used for special wheelchair pattern, by changing the form of operation, can be wheelchair occupant, and safe use is provided.
As the wheelchair pattern, have situations such as following, this situation refers to, such as, make the rising or falling speed of cab low than usual, in addition, when stopping at every floor plate, it is high at ordinary times to make floor between cab and the Room floor arrive ratio of precision.Take advantage of the occasion of seat normal health person, rising or falling speed is reduced, in addition, do not improve floor especially and arrive precision, thereby can reduce the traveling time of elevator, and then operational efficiency is improved.
For according to such, passenger's kind automatically changes the operational mode of elevator, must automatically discern the situation whether wheelchair occupant plans to take advantage of seat.
In the past, invent according to this purpose, the image-processing system of implementing is non-existent, wheelchair occupant is operated the button of wheelchair occupant special use in person, or is entered in the cab when calling out cab, same wheelchair button is operated, thus, must provide the switching indication of operational mode to elevator device.
In addition, with regard to the pattern recognition of elevator passenger, such as, the flat 6-92563 document of TOHKEMY discloses the number determinator of waiting Passengen.Figure 16 and Figure 17 are the figure of the content of the technology that is used to illustrate that the document is put down in writing.
Figure 16 discloses the composition of number determinator in the past.In the figure, this installs by video camera 1001, image-processing system 1002, and neural network 1 003, number judgment means 1004, conveyer 1005, elevator control gear 1006 forms.Television camera be arranged at be positioned at waiting Passengen the head on smallpox board place, vertically towards.
At first, by television camera, take image.In the input imagery 1010 of Figure 17, as an example, provide the appearance that is decorated with 3 people.At the state that does not have the people, prepare the image of having taken in advance, promptly background images 1011, and input imagery and this background images are carried out difference processing, and thus, difference appears in people's zone.The example of difference image is by 1012 expressions of the label among Figure 17.This image is carried out double quantification handle, occur the people zone clearly.Among this figure 1013 expression double quantification result.This a series of processing is undertaken by image-processing system 1002.
Then, binary picture is divided into 12 pieces, is entered in the neural network 1 003.By the neural network of learning in advance, above-mentioned input data are handled, by number judgment means 1004, output data is handled, number is judged.This result sends elevator control gear 1006 to by conveyer 1005.
In the image-processing system that provides, produce following phenomenon in the above, that is, in the situation of seeing a plurality of passengers, can't separate clearly and the corresponding corresponding image region of passenger with overlapping, even under the situation of binary images, a plurality of passengers are equivalent to 1 zone.In prior art,, video camera is gone up setting towards people's head downwards for fear of this phenomenon, but owing to have following problems, that is, and in the elevator hall, video camera is installed on meeting cost cost on the smallpox board, thus require video camera is arranged within the elevator cage, or on the outer frame of elevator door.That is, be provided with in the condition, because passenger's zone can't be separated on image accordingly, so have the pattern recognition problem of failure easily at general video camera.
In addition, for the people is discerned, such as, disclose the method for extracting the face zone in the TOHKEMY 2000-193420 document.In the document, point out feature according to the bridge of the nose, detect the position of eye, still, can't judge people's state, i.e. the state that has or not of wheelchair.
Summary of the invention
The objective of the invention is in order to solve above-mentioned existing problem, the image-processing system of a kind of people's who judges shooting state is provided, specifically, provide identification automatically to take advantage of the passenger of an elevator whether to take advantage of the image-processing system of seat on wheelchair, and the elevator that is mounted with this image-processing system.
In the present invention, at first from the image of taking, extract the face zone, then,, extract supposition, judge have or not (state) of wheelchair such as, the 2nd zone of wheelchair from this zone.
Image-processing system of the present invention comprises: face region extracting device, this face region extracting device extract the 1st zone of the face that comprises the personage from the personage's that taken image; The surveyed area estimating device, this surveyed area estimating device adopts the zone of extracting by above-mentioned face region extracting device, infers in the above-mentioned image, in advance Deng Ji required surveyed area; Judgment means, this judgment means are judged the image in the 2nd zone of extracting according to this surveyed area estimating device, judge whether to meet the condition of registration in advance.According to this scheme, according to the face zone, people's corresponding judgement can correctly be carried out, and according to this information, can extract the zone of the information that people's plan understands apace, can obtain people's information.
In addition, the image-processing system of another invention comprises: television camera, and this television camera is provided with dividing outdoor mode of taking according near the door that will divide indoor and outdoor; Face region extracting device, this face region extracting device detect the position in personage's face zone according to the image by above-mentioned shot with television camera; Wheelchair zone estimating device, this wheelchair zone estimating device is taken advantage of the occasion of seat on wheelchair the above-mentioned personage of supposition, adopts the above-mentioned face regional location that has extracted, according to above-mentioned image, infers the zone that is equivalent to the wheelchair part; Wheelchair has or not judgment means, and this wheelchair has or not judgment means according to the picture information in the above-mentioned wheelchair zone of having inferred, judges having or not of wheelchair existence.According to this scheme, according to the face zone, infer passenger's position, according to this situation, infer wheelchair should the position, carry out image processing, thus, limit image region, at a high speed and positively carry out the judgement that wheelchair exists.
In addition, this device also comprises Distnace determination device, this Distnace determination device is used to be measured to the distance in the face zone of extracting by above-mentioned face region extracting device, in the estimating device of wheelchair zone, adopt the above-mentioned range data of having measured, infer the wheelchair zone, thus, wheelchair exists the precision of the judgement that has or not to improve.That is, when the position of the face that detects the passenger, pass through Distnace determination device, infer the position in 3 dimension spaces of this passenger face position, thus, can calculate the height of face's relative floor in position of this passenger, if it can judge clearly that then this passenger does not take advantage of a wheelchair outside predetermined scope.If in this predetermined scope, even then take advantage of the occasion of seat on wheelchair the passenger, can be on image, infer the zone at wheelchair place, can have or not judgment means by wheelchair, there is intra-zone analysis in above-mentioned wheelchair of having inferred, whether the decidable wheelchair exists.Therefore, the precision of judgement improves.
Also have, above-mentioned Distnace determination device comprises optical beam generating device, the light beam of this optical beam generating device origination point or seam shape; Light-beam scanner, this light-beam scanner scans according to the mode that the light beam of above-mentioned generation projects determined zone; The reflected light pick-up unit, this reflected light pick-up unit detects the reflected light of the light beam that throws through scanning, according to above-mentioned beam flying angle and reflected light detection position, detects the distance of object, thus, can carry out range determination simply.
Have, said apparatus comprises the range gate setting device again, and this range gate setting device is set and is used to judge the approaching at a certain distance distance threshold of dividing the situation of chamber of personage; The picture-taking timing device, this picture-taking timing device is discerned the situation that the personage enters in the range gate by Distnace determination device, determines to take the moment of image; According to the above-mentioned fixed moment, by television camera, image is taken, thus, effectively, and positively obtain image.In addition, can be at passenger near video camera, keep certain moment substantially in the distance of distance video camera, take image, thus, can on image,, the reliability of pattern recognition be improved according to the face and the wheelchair of certain substantially size taken of passengers.
Elevator of the present invention is mounted with above-mentioned any one described image-processing system, television camera is near the door of above-mentioned elevator, the television camera that the passenger is taken, to have or not the information of judgment means acquisition by wheelchair, be input to above-mentioned elevator control device, thus, following elevator can be provided, the corresponding passenger of this elevator and change the running status of elevator realizes more comfortable operation, the performance height.
Description of drawings
Fig. 1, Fig. 2 are the block scheme of composition that is used to illustrate the image-processing system of the 1st embodiment of the present invention, and Fig. 3 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Fig. 4 is the synoptic diagram that is used to illustrate the image-processing system of the 2nd embodiment of the present invention, the figure shows 3 passengers take advantage of seat from the foreign side of elevator appearance.Fig. 5 is by the synoptic diagram of the object image of the image-processing system processing of the 2nd embodiment, wherein the passenger among Fig. 4 is taken; Fig. 6 is the block scheme of composition that is used to illustrate the image-processing system of the 2nd embodiment, and Fig. 7 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Fig. 8 is the block scheme of composition that is used to illustrate the image-processing system of the 3rd embodiment of the present invention, and Fig. 9 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Figure 10 is the block scheme of composition that is used to illustrate the image-processing system of the 4th embodiment of the present invention, and Figure 11 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Figure 12 is the block scheme of composition that is used to illustrate the image-processing system of the 5th embodiment of the present invention, and Figure 13 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Figure 14 is the block scheme of composition that is used to illustrate the image-processing system of another embodiment of the present invention, and Figure 15 is the process flow diagram of the flow process of the action of this image-processing system of expression;
Figure 16, Figure 17 is for representing in the past the composition of the number determinator relevant with the Figure recognition of elevator passenger, block scheme and the figure that is used to illustrate treatment step.
Embodiment
Be used to realize preferred form of the present invention
Below with reference to the accompanying drawings, embodiments of the invention are described.
The 1st embodiment
Fig. 1, Fig. 2 are the figure of composition that is used to illustrate the image-processing system of 1 embodiment of the present invention, and Fig. 3 is the process flow diagram of the flow process of the action of this image-processing system of expression.
In the drawings, label 1 expression such as, the filming apparatus that digital camera is such, label 2 expressions are from the visual 2a that has taken, extract the face region extracting device of face zone 2b, label 3 expression surveyed area estimating devices, this surveyed area estimating device is according to the part of the face zone 3b (being equivalent to 2b) of the visual 3a (being equivalent to 2a) that has taken, infer the surveyed area 3c of registration in advance, the device that label 4 expression compares the condition of the picture information of 3c and registration in advance, label 5 is represented the output of these judgment means.
Below according to the process flow diagram of Fig. 3, the step of image processing is described.
At first, in step 101 (in the accompanying drawing in the back, being called ST101), from television camera, the input gray level image.Should take image storage in storer.Then, in step 102,, extract the face zone 2b of the people among the visual 2a by face region extracting device 2.As people's face extracted region algorithm, adopt such as, based on the mode of the extraction of the skin color part of color image.Be positioned at the occasion of visual field a plurality of people, extract a plurality of zones, its output.Then, in above-mentioned face region extracting device 2,, measure the position in continuous face zone in step 103.In the present embodiment, with such as, the value of visual center of gravity in zone is as the representative point position in zone.
Then, in step 104,,, carry out the zone of required surveyed area and infer according to the above-mentioned face regional location that has extracted by surveyed area estimating device 3.In Fig. 1, be the extraction in car chair zone, at this occasion, the example of the presuming method in car chair zone is specifically described.Measure the face zone of extracting in the above-mentioned face region extracting device.With with the relevant reference data of transverse width of people's general face, Ying Yuqi compares, and thus, can infer this face zone distance of video camera relatively.Video camera the angle of depression is set, be that angle between camera lens optical axis and the horizontal line is measured when video camera is provided with, it is as can be known like this, thus, adopt above-mentioned face the zone distance and the video camera angle of depression of video camera relatively, infer the above-mentioned face zone height of floor relatively.In addition, can infer the face center of the position of wheelchair, the below that keeps certain distance substantially the passenger who rides relatively.Can adopt these data, can there be the zone on the image of wheelchair in setting.
Then,, search for above-mentioned wheelchair zone of having inferred, judge having or not of wheelchair in step 105.Wheelchair detects and to be undertaken by following manner, and this mode is: such as, wheelchair adopts the seat portion of passenger institute seat and is positioned at the geometric feature of two wheels of its both sides, colouring information etc.These information can be used as the data of registration in advance, compare judgement.
At last, be in the above-mentioned wheelchair judgment means, judge that wheelchair is positioned at the occasion in zone, judge that this passenger rides, produce the wheelchair detection signal, a series of processing is all over.
In above-mentioned example, judge having or not of wheelchair shown in Figure 1, still, it is suitable for equally with safety is purpose, such as, as shown in Figure 2, the judgement that has or not etc. of attaching the names of pre-determined candidates.In Fig. 2, can further analyze the top among the face zone 2b (3b), compare with the inferring of shape of the general head of the state that does not have cap, thus, and can be according to colouring information, the judgement that has or not of attaching the names of pre-determined candidates.
In addition, in the above-described embodiments, in the television camera visual field, the occasion of only extracting 1 face zone is described, but, in occasion, similarly with a plurality of faces zone, at corresponding face zone, carry out step 102 repeatedly～105 corresponding steps, can judge thus.
In addition, in the present embodiment, the face regional location adopts the regional barycenter position, still, also can adopt the external tetragonal center in face zone, or, detect the eye in the face zone according to concrete image processing, the part of the feature of nose etc., according to them, the calculation representative point position.
Also have, in the present embodiment, to as the face region extracting device, method based on the skin-color extracted region that adopts color image is described, but obviously, even, under the situation that realizes extracting, also can obtain the effect identical with present embodiment in the identical skill and technique of the pattern of the face pattern that comprises face by detection.
Have again, in the present embodiment, above-mentioned wheelchair judgment means is with according to geometric feature, the method of judging the existence of wheelchair is that an example is described, but obviously,, can distinguish the further feature in wheelchair occupant and walking normal health person's the wheelchair zone even take advantage of the state etc. of the passenger of seat on wheelchair shank in employing, under the situation about judging, still can implement present embodiment.
As present embodiment, from the image that comprises the people, at first extract people's face zone, then, according to the information in face zone, infer, detect required surveyed area, can judge having or not of wheelchair, the situation that has or not etc. of attaching the names of pre-determined candidates, thus, from image, only cut out necessary information successively, the speed of processing improves, and final judgement precision improves.
The 2nd embodiment
Below another embodiment of the present invention is described.Fig. 4 is used to illustrate the image-processing system of one embodiment of the present of invention, and this figure represents the foreign side of passenger from elevator in a schematic way from the side, takes advantage of the state of seat.The object image that Fig. 5 carries out for the image-processing system by present embodiment, it is a synoptic diagram of taking the state of Fig. 4 by video camera.Fig. 6 is the block scheme of the composition of the image-processing system of expression present embodiment, and Fig. 7 is the process flow diagram of the flow process of the action of this image-processing system of expression.
In the drawings, label 11 expressions are arranged at the television camera of the external door top of elevator 16, and to taking advantage of the passenger 17a of an elevator, 17b takes.The passenger, such as, for utilizing the passenger 17a of wheelchair, do not utilize the passenger 17b of wheelchair.Label 12 expression color region extraction elements, label 13 expression wheelchair zone estimating devices, label 14 expression wheelchairs have or not judgment means, and label 15 expression wheelchairs have or not detection signal.
Below according to the process flow diagram of Fig. 7, the action of this image-processing system is described.In step 201, the gray scale image that input is taken by television camera 11.This television camera is arranged at the outer arch of external door, its sight line towards around look around the such direction of elevator waiting Passengen.With the image storage of having taken in video memory.This image, that is, and the image of the shot with television camera by Fig. 4, such as, for shown in Figure 5.
Then, in step 202,, extract the face zone of the people in the image by face region extracting device 12.As people's face extracted region algorithm, adopt method based on the skin color extraction partly of color image.Be positioned at the occasion of visual field a plurality of people, extract a plurality of zones, its output.It is equivalent among Fig. 5, the hachure zone in the corresponding zone that is illustrated by the broken lines.Then, in step 203, measure the position in continuous face zone.In the present embodiment, identical with the 1st embodiment, with the value of the visual center of gravity in zone, as the representative point position in zone.
Then,, adopt wheelchair zone estimating device 13 in step 204, according to above-mentioned face zone of having extracted, if infer this people's seat on this wheelchair, the position of the wheelchair of Cun Zaiing then.This situation depends on following method, and is promptly described as the foregoing description, and the transverse width in the face zone of extracting in above-mentioned face region extracting device is measured etc.Reference data by with the transverse width of people's general face compares with it, can infer the distance of this face zone and video camera.This video camera the angle of depression is set, be that angle between camera lens optical axis and the horizontal line is measured when video camera is set, it may be known thus, adopts the distance and the video camera angle of depression of above-mentioned face zone and video camera like this, infers the height of above-mentioned face zone and floor.In addition, if the position of wheelchair is positioned at the face center with the passenger that sat, the below that keeps certain distance substantially then can be inferred, and like this, can adopt these data, sets the zone on the image that wheelchair can exist.
Then,, have or not judgment means 14, search above-mentioned wheelchair zone of having inferred, judge having or not of wheelchair by wheelchair in step 205.The detection of wheelchair by such as, following manner is carried out, this mode is: wheelchair adopts the seat portion that the passenger sat and is positioned at the geometric feature of two wheels of its both sides, colouring information etc.
Below, in the above description, adopt the wheelchair of gray scale image pattern (pattern) to have or not judgement to specifically describe especially.
In order to judge having or not of wheelchair, at first, make mode of learning (pattern) in advance, be equipped with a plurality of study data.Should " study data " refer in image, make and take the pictorial data that wheelchair is arranged, passenger's clothes, the kind of wheelchair, various variations take place in the moment of shooting, prepare a plurality of data.According to certain rule (such as, raster scanning), replace from these study with what cut out the image by 1 dimensional signal, the image of the shank near zone of a plurality of wheelchair occupant on the hyperspace of the brightness value that each is visual as the value of each, is drawn.Make by being equivalent to the zone that these a plurality of wheelchairs " have " visual graphical pointv group to form, forming wheelchair " has " study point group.
To adopting this mode of learning, judge that the step that has or not of wheelchair is described below.From unknown image,, there is the zone of wheelchair according to the information of above-mentioned face image region.By generating identical method, in hyperspace, draw the image region that has extracted with above-mentioned learning data.It is the unknown number strong point.Then, calculate unknown data respectively, and form the distance between the respective point of above-mentioned a plurality of study point groups, pre-set threshold is compared.If have the learning data point of the unknown data distance following, judge that then this unknown data is that wheelchair " has " with being positioned at threshold value.
At last,,, judge that judge that passenger seat on wheelchair, produces the wheelchair detection signal, a series of processing is all in the occasion of region memory at wheelchair by above-mentioned wheelchair judgment means in step 206.
Here, in the present embodiment, to in the television camera visual field, the occasion of extracting 1 face zone is described, still, can be as shown in Figure 5, in the occasion that has a plurality of faces zone, similarly to corresponding face zone, carry out step 202～206 corresponding steps, thus, can confirm whether be respectively located on the wheelchair near a plurality of passengers of elevator cage.
In addition, described as above-mentioned the 1st embodiment, the face regional location adopts the external tetragonal center in face zone, also can detect the eye in face zone, the characteristic of nose etc. in addition by concrete image processing, according to this testing result, the position of calculation representative point.
Also have, described as above-mentioned the 1st embodiment, obviously, as the face region extracting device, even by the input soot-and-whitewash, detect under the situation of the situation that method that the pattern of the face pattern of the face comprise nose etc. coincide proposes, still can obtain the effect identical with present embodiment.
Have again, described as above-mentioned the 1st embodiment, as above-mentioned wheelchair judgment means, to judge that according to geometric feature the method that exists of wheelchair is that an example is described, still, obviously, can adopt the situation of the shank of taking advantage of the passenger of seat on wheelchair etc. equally, further feature in difference wheelchair occupant and walking normal health person's the wheelchair zone is judged, realizes present embodiment.
Described as present embodiment, owing to arrive the arrival floor floor at elevator cage, in the elevator hall, wait for that the passenger of elevator cage enters in the interior process of elevator cage, can not have under other people the hidden state, by obtaining its image corporally, so can realize stable pattern recognition.In addition, have following effect, that is, and in identification, according to as being difficult to face zone by other people hidden higher zone, the position of inferring the passenger, inferring wheelchair in addition should be in the position, qualification image processing zone, thus, wheelchair exists the judgement that has or not stably to carry out.
The 3rd embodiment
Below another embodiment of the present invention is described.Present embodiment is according in above-mentioned the 2nd embodiment, and except having the video camera of making a video recording, the mode of also adding the Distnace determination device of object image to forms.
Fig. 8 is the block scheme of the scheme of the image-processing system of expression present embodiment, and Fig. 9 is the process flow diagram of the flow process of the action of this device of expression.
In the figure, label 11～15 is identical with the foregoing description 2, label 18 expression Distnace determination devices.In addition, the Distnace determination device in the present embodiment is for the television camera that adopts eyes, and the example of the method that stereoscopic image is resolved is though this point is not shown in the drawings.
According to Fig. 9, the order of moving is described below.In addition, because step 301～303 are identical with above-mentioned the 1st embodiment, so be described briefly.At first,,, extract the face zone,,, detect the face regional location according to gravity model appoach etc. in step 303 in step 302 according to the image of television camera input.In addition, owing to be the bimirror video camera, the image of the corresponding video camera of basis carries out the processing of above-mentioned steps 301～303 respectively.
Then, in step 304,, measure distance by Distnace determination device 18.In the present embodiment, as method for measuring distance, be fit to adopt the bimirror video camera.That is,, set in advance 2 known television cameras of geometric position relation, obtain the corresponding relation between each pixel of image that obtains by each video camera as television camera, thus, according to the principle of triangulation, the range distribution of determination object thing.This range distribution is stored in each pixel as the distance image of having stored the distance from the video camera to the object.
Then, in step 305, by wheelchair zone estimating device 13, according to the position on the image in above-mentioned face zone of having extracted, and corresponding with this face zone, the range determination result who has measured calculates the face zone height of floor relatively.In the present embodiment, for the gray scale image that obtains by television camera,, by calibration operation, obtain locational correspondence in advance with above-mentioned distance image.Here, identical with above-mentioned the 2nd embodiment, the angle of depression of television camera is known, thus, by will be comprehensive with it to the range data in face zone from above-mentioned video camera, assert the position of 3 dimension spaces in face zone.Therefore, the height to the relative floor in face zone calculates.
Afterwards, in step 406,, regional height and the pre-set threshold to floor of this face of having measured compared by this wheelchair zone estimating device 13.Threshold setting in the present embodiment is the height of the face of common adult when standing, the average height of the face when taking advantage of seat on wheelchair.,, judge the zone of this face zone here, detect and handle if carry out wheelchair, then end process for upright pedestrian in the height of floor from rail top in face zone occasion greater than this threshold value.But the height of floor is less than the occasion of this threshold value relatively in the face zone, and the possibility in the zone that this face zone is a wheelchair occupant is higher, infers the step 307 of the position on the image that has the wheelchair zone.
In step 308 and step 309, have or not judgment means 14 by wheelchair, specifically investigate above-mentioned wheelchair zone of having inferred, judge having or not of wheelchair, the wheelchair detection signal takes place in the occasion in that judgement has.These actions respectively with the step 205 of the 2nd embodiment, identical with 206.
Here, in the present embodiment, as Distnace determination device 18, adopt the bimirror anaglyph, but, even adopt the above video camera method for measuring distance of three mirrors in the method for the distance that adopts single mirror video camera to detect from the focusing blurred image etc., have and measure above-mentioned face zone of having extracted relatively under other the situation of method for measuring distance of the function of the distance of video camera, still obtain identical effect.
In addition, in the present embodiment, provided just to make and comprised the distance image, it is the example of data structure of form of the range information of the relative video camera of each pixel, but as above-mentioned range data, can exemplify the corresponding data in representative point position with above-mentioned face zone of having extracted, even in the following cases, still obtain same effect, this situation refer to such as, the visual field of above-mentioned television camera is divided into polylith, at every pars intermedia, the range determination sensor of some rayed mode of the distance of the relative video camera of a plurality of mensuration is set, or the narrower ultrasonic sensor of a plurality of directive property is set.
Described as present embodiment, by the passenger's that judges rightly 3 locations of tieing up, can correctly carry out inferring of wheelchair zone, consequently, have to make and judge the effect of this passenger for the reliability raising of wheelchair occupant.
The 4th embodiment
Below another embodiment of the present invention is described.
Figure 10 is the block scheme of the composition of the image-processing system of expression present embodiment, and Figure 11 is the process flow diagram of the part of the flow process of the action of this device of expression.This embodiment 4 relates to following image-processing system, and this image-processing system is characterised in that the Distnace determination device of the 3rd embodiment adopts laser range sensor.
In the figure, label 11～15 and the foregoing description 2,3 is identical.Label 19 expression range determination control device, label 20 expression optical beam generating devices, label 21 expression light-beam scanners, label 22 expression reflected light determinators.
According to the process flow diagram of Figure 11, action is described below.In addition, Figure 11 is another embodiment of the step 304 among Fig. 9, and " beginning " among Figure 11 is arranged in the back of the step 303 of Fig. 9, and " end " among Figure 11 is connected on step 305 back among Fig. 9.At first, by television camera 11, input imagery by face extraction element 12, from image, extracts people's face zone, in addition, measures the position that the face Regional Representative is ordered.
Then, carry out the later range determination of step 401.This range determination control device 19, optical beam generating device 20, light-beam scanner 21, the device that carries out the control of reflected light pick-up unit 22.The principle of the range determination that present embodiment provides is to send the slit-type laser beam, and by video camera, the light that returns running into object is taken, and according to the principle of triangulation, measures distance, and it is called " light cross-section method ".Optical beam generating device 20 is formed by semiconductor laser generating device and lens, and it is a following apparatus, and the output laser beam that will encourage is for the time being assembled, and in addition by lens pillar, produces the light that is the slit-shaped distortion.In the initial step 401 of range determination,, the luminous pattern of laser beam is carried out initialization process by above-mentioned range determination control device 19.Luminous pattern refers to when laser slit light is scanned, and illuminates the pattern that extinguishes synchronously with scanister.In the present embodiment, in the following manner, in the scanning of 1 width of cloth picture, only illuminate in the time, the luminous pattern that extinguishes in other interval carries out initialization process, and this mode is: generation makes the moment of illuminating carry out a plurality of patterns of various variations.
Then, in step 402, the control pattern of 21 pairs of laser beam flyings of light-beam scanner carries out initialization process.This means according to above-mentioned light-beam scanner and image and obtain the mode that constantly synchronously scans, to sweep velocity, scope is carried out initialization process.
Afterwards, in step 403, beginning is as the image storage of the ccd video camera of reflected light pick-up unit 22.In step 404, according to the pattern that has carried out above-mentioned initialization process, begin action as the Electronic Speculum of light-beam scanner 21, in step 405, optical beam generating device 20 is according to the pattern that has carried out above-mentioned initialization process, and the ON/OFF of light beam is controlled.In step 406, whether laser beam flying is finished to judge.If finish, then carry out step 404 repeatedly～405 action, if finish,, the image storage of reflected light pick-up unit 22 is finished then at next step 407.
Then,, read the laser beam reflected light image of above-mentioned storage,, measure its position by image processing in step 408.Position according to the beam reflection light on the image that has detected, and this laser beam is in the scanning angle of light-beam scanner 21 in the moment of ON, principle based on triangulation, measure 3 dimension positions of the position on the target object that this light beam throws, promptly, distance in step 409, is stored in range data in the storer in the range determination control device 19.
In the present embodiment, owing to only shine 1 time mode according to the seam shape laser beam that in memory period, is scanned as every images of the ccd video camera of reflected light pick-up unit, set luminous pattern, so can in every width of cloth memory image, measure the range distribution of the perpendicular every row of direction.Therefore, by when making above-mentioned light beam be the moment variation of ON, carry out step 403 repeatedly～409, then can be to all places, illuminating laser beam, the distance of mensuration irradiation position.Consequently, in step 409, obtain the distance image.As mentioned above, the step of range determination finishes.
The continuous action of present embodiment is the such action shown in step 305～309 among Fig. 9, promptly, by wheelchair zone estimating device 13, according to the position on the image in the face zone of having extracted and corresponding with this face zone, the range determination result who has measured calculates the face zone height of floor relatively.By this wheelchair zone estimating device 13, with the regional height of floor relatively of this face of having measured, compare with pre-set threshold, carry out the roughly judgement that wheelchair has or not.The height of floor is less than the occasion of threshold value relatively in the face zone, and the possibility in the face zone that this face zone is a wheelchair occupant is higher, infers the position on the existing image in wheelchair zone.At last,, carefully analyze above-mentioned wheelchair zone of having inferred, judge having or not of wheelchair,, the wheelchair detection signal takes place in the occasion of judging " having " by this wheelchair zone estimating device 13.
In addition, in the present embodiment, as by optical beam generating device, light-beam scanner, the Distnace determination device that the reflected light pick-up unit forms, to when the luminous moment is changed, at random carry out repeatedly in memory period as every images of the ccd video camera of reflected light determinator, produce the light cross-section method of 1 laser slit shape bundle, the method that generates the distance image is described, even but be used in every images memory period of above-mentioned ccd video camera, throw under the situation of space encoding facture of a plurality of laser slit shape bundles, still obtain same effect.
In addition, in the present embodiment, provided light-beam scanner 21 and be fit to adopt the example of Electronic Speculum, but also can adopt the polygon mirror of polygon prism type.
Described as present embodiment owing to adopt the Distnace determination device of projecting beam, so obtain following effect, that is, and can not rely on the clothes that the passenger wears pattern have or not with complicacy, stable range determination.
The 5th embodiment
Below an also embodiment of the present invention is described.
Figure 12 is the block scheme of the composition of the image-processing system of expression present embodiment, and Figure 13 is the process flow diagram of the flow process of the action of this device of expression.
In the drawings, label 11～15,18 identical with the 2nd～4 embodiment.Label 23 presentation images are taken timing device, label 24 expression range gate setting devices.
According to the process flow diagram of Figure 13, present embodiment is described below.
At first, in step 501, by range gate setting device 24, the setpoint distance door.This range gate refers to the occasion in the elevator cage inside of opening door near the passenger, the threshold value of the distance between television camera 11 and the passenger, such as, for passing through above-mentioned television camera 11, investigate in advance according to the face zone of required size taken of passengers necessary, the video camera of inferring and the threshold value of the distance between the passenger.
Then, in step 502,, measure gamma camera and passenger's distance by Distnace determination device 18.This Distnace determination device be fit to adopt such as, above-mentioned the 3rd embodiment, or the scheme of the 4th embodiment.
Afterwards, in step 503, with above-mentioned video camera of having measured and the distance between the passenger, compare with the above-mentioned range gate threshold value of having set, if the distance between video camera and the passenger as yet near and be not positioned in the range gate, then once more by step 502, measure approaching passenger's distance.Distance between video camera and passenger near and be positioned in the range gate, then in step 504,, determine to take the moment of image by picture-taking timing device 23.Wherein, in the image after shooting, be similar to the mode of predefined value, determine this moment according to the size in the people's who has taken face zone.
Then, in step 505, by television camera 11, the image of taken of passengers in step 506, by face region extracting device 12, extracts the face zone, and by step 507, detects the position of the representative point in above-mentioned face zone.In addition, in step 508,, infer and think the zone that has wheelchair according to the position in the face zone of having extracted by wheelchair zone estimating device 13.The action of wheelchair zone estimating device such as, provide among the 2nd embodiment.
In addition,, have or not decision maker 14, in above-mentioned wheelchair zone of having inferred, carry out image processing, judge having or not of wheelchair by wheelchair in step 509.Consequently, in the occasion of judging " having ", in step 510, output wheelchair detection signal 15, a series of action is all over.
Here, in the present embodiment, as Distnace determination device, to the 2nd, or the mode that provides among the 3rd embodiment is described, but can be the sensor with following function, this function refers to find the relative video camera of passenger, near and in certain distance, obviously, such as, even employing photoelectric sensor, ultrasonic sensor under the situation of electrothermic type people induction pick-up etc., still obtains same effect.
In addition, in the present embodiment, to output with Distnace determination device, only the method that compares with range gate is described, if but Distnace determination device is fit to the device that employing can obtain the distance image, then following situation also can realize by the device of the composition identical with the scheme of Figure 12, this situation refers to as the 2nd inventive embodiment described, by calculating the regional distance of having extracted by the face region extracting device of video camera relatively of face, move according to the mode that improves the reliability of inferring in the wheelchair zone.
Described as present embodiment, because can be at passenger near video camera, distance at its relative video camera keeps certain moment basically, take image, thus following effect had, promptly, can be on image, according to certain basically size, the face of taken of passengers and wheelchair can make the reliability of pattern recognition improve.
The 6th embodiment
Below embodiments of the invention are described.Present embodiment relates to the image-processing system that is mounted with above-mentioned the 2nd～5 embodiment, in the above-described embodiments, with the wheelchair exported have or not the judgement signal, be input to elevator control device, the run action of elevator is controlled.
Figure 14 is the block scheme of composition that is used to illustrate the image-processing system of present embodiment, and Fig. 9 is the process flow diagram of flow process of the action of presentation image treating apparatus.
In the figure, label 11～15 is same as the previously described embodiments, label 25 expression elevator control gears.
Wheelchair identification signal by the image processing of adopting by above-mentioned the 2nd～5 embodiment obtains is entered in the following elevator control gear.As control action, provide following such action by example.
(a) opening and closing speed of elevator door (, realizing at leisure opening and closing) to taking advantage of the people of a wheelchair;
(b) the opening the time of elevator door (in the people's who takes advantage of a wheelchair occasion, door can be opened for a long time);
(c) adjustment of elevator floor (for the control of the discrepancy in elevation between the floor face in the cab of elevator and the outside, when the people of wheelchair takes advantage of, can accurately make above-mentioned both be consistent, can not produce the discrepancy in elevation);
(d) guiding of sound (sounding of broadcast " operating in now in the accessible pattern " etc.);
(e) an elevator translational speed (when the people of wheelchair takes advantage of, slowing down the mobile maximal rate and the acceleration of elevator);
(f) when the button control of elevator inside (when the people of wheelchair takes advantage of, make the wheelchair dedicated button that is arranged at the lower position in the cab effective, when not being such, make it invalid, or can be only push for a long time continuously etc. push the time effective);
(g) control of the special sensor in the elevator cage is (when the people of wheelchair takes advantage of, make disabled person's sensor special of the voice recognition that is arranged in the cab etc. effective), by carrying out above-mentioned control, can realize the operation of the preferred elevator of wheelchair occupant, operational efficiency is improved, and can be the passenger comfortable operation is provided.
In addition, Figure 14 of the foregoing description mainly represents forming of the image-processing system that provided by the 2nd embodiment, but obviously, also can be and comprise the Distnace determination device that provides by the 3rd embodiment, the range determination control device that provides by the 4th embodiment, optical beam generating device, light-beam scanner, reflected light pick-up unit, the Distnace determination device that provides by the 5th embodiment, the range gate setting device, the scheme of the image-processing system of picture-taking timing device.
Image-processing system of the present invention is used for automatically detecting the passenger who plans to take advantage of an elevator, comprises the occasions such as situation of taking advantage of the passenger of seat on wheelchair, corresponding to this situation, is used to change the control of the operational mode etc. of elevator.
Claims (6)
1. image-processing system, this image-processing system comprises: face region extracting device, this face region extracting device extract the 1st zone of the face that comprises the personage from the personage's of picked-up image; The required surveyed area of logining is in advance inferred in the above-mentioned image in the zone that surveyed area estimating device, this surveyed area estimating device adopt above-mentioned face region extracting device to extract; Judgment means, this judgment means are judged the image in the 2nd zone of extracting according to this surveyed area estimating device, judge whether to meet the condition of registration in advance.
2. image-processing system, this image-processing system comprises: television camera, this television camera is provided with dividing outdoor mode of taking according near the door that will divide indoor and outdoor; Face region extracting device, this face region extracting device detect the position in personage's face zone according to the image by above-mentioned shot with television camera; Wheelchair zone estimating device, this wheelchair zone estimating device is taken advantage of the occasion of seat on wheelchair the above-mentioned personage of supposition, adopts the above-mentioned face regional location that has extracted, according to above-mentioned image, infers the zone that is equivalent to the wheelchair part; Wheelchair has or not judgment means, and this wheelchair has or not judgment means according to the picture information in the above-mentioned wheelchair zone of having inferred, judges having or not of wheelchair existence.
3. image-processing system according to claim 2, it is characterized in that it also comprises Distnace determination device, this Distnace determination device is used to be measured to the distance in the face zone of extracting by above-mentioned face region extracting device, in the estimating device of wheelchair zone, adopt the above-mentioned range data of having measured, infer the wheelchair zone.
4. image-processing system according to claim 3 is characterized in that above-mentioned Distnace determination device comprises optical beam generating device, this optical beam generating device generation point-like, or the light beam of seam shape; Light-beam scanner, the mode that this light-beam scanner makes the light beam of above-mentioned generation project determined zone scans; Reflected light pick-up unit, this reflected light pick-up unit detect the reflected light of the light beam of scanning projection,
According to above-mentioned beam flying angle and reflected light detection position, detect the distance of object.
5. image-processing system according to claim 3 is characterized in that comprising: range gate setting device, this range gate setting device are set and are used to judge the approaching at a certain distance distance threshold of dividing the situation of chamber of personage; The picture-taking timing device, this picture-taking timing device is discerned the situation that the personage enters in the range gate by Distnace determination device, determines to take the moment of image; According to above-mentioned definite timing,, image is taken by television camera.
6. elevator, this elevator is equipped with any one the described image-processing system in the claim 2～5, it is characterized in that television camera is near the door of above-mentioned elevator, the television camera that the passenger is taken, to have or not the information of judgment means acquisition by wheelchair, be input to above-mentioned elevator control device.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2000/009319 WO2002056251A1 (en) | 2000-12-27 | 2000-12-27 | Image processing device and elevator mounting it thereon |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1434959A true CN1434959A (en) | 2003-08-06 |
| CN1261904C CN1261904C (en) | 2006-06-28 |
Family
ID=34140839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN00819023.2A Expired - Fee Related CN1261904C (en) | 2000-12-27 | 2000-12-27 | Image processing device and elevator mounting it thereon |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP4476546B2 (en) |
| CN (1) | CN1261904C (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102292277A (en) * | 2008-12-11 | 2011-12-21 | 因温特奥股份公司 | Method for enabling the use of an elevator system by disabled persons |
| CN101263074B (en) * | 2005-09-16 | 2013-03-27 | 奥蒂斯电梯公司 | Optically monitoring comb-line of escalators and moving walks |
| CN104743419A (en) * | 2013-12-25 | 2015-07-01 | 株式会社日立制作所 | Image monitoring device and elevator monitoring device |
| CN105989343A (en) * | 2014-09-25 | 2016-10-05 | 株式会社日立建筑系统 | Detection device and method of seeing-eye dog |
| CN111362081A (en) * | 2020-04-15 | 2020-07-03 | 上海三菱电梯有限公司 | Elevator group management method and elevator group management system |
| CN111492371A (en) * | 2017-12-14 | 2020-08-04 | 三菱电机株式会社 | Retrieval system and monitoring system |
| CN111507125A (en) * | 2019-01-30 | 2020-08-07 | 佳能株式会社 | Detection device and method, image processing device and system |
| CN113939469A (en) * | 2019-06-20 | 2022-01-14 | 三菱电机株式会社 | Control device for elevator |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018235279A1 (en) * | 2017-06-23 | 2018-12-27 | 日本電気株式会社 | Information processing device, control method, and program |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04116085A (en) * | 1990-09-03 | 1992-04-16 | Mitsubishi Electric Corp | Elevator applicable for wheel chair |
| JPH05147835A (en) * | 1991-11-25 | 1993-06-15 | Hitachi Ltd | Elevator stop calling register for disabled people |
| JPH061546A (en) * | 1992-06-17 | 1994-01-11 | Hitachi Ltd | Elevator control device with operational specification for wheel chair |
| JP3293308B2 (en) * | 1994-03-10 | 2002-06-17 | 三菱電機株式会社 | Person state detection device |
| JPH09198508A (en) * | 1996-01-17 | 1997-07-31 | Mitsubishi Electric Corp | Eye state detector |
| JPH11268879A (en) * | 1998-03-20 | 1999-10-05 | Mitsubishi Electric Corp | Operation controller for elevator |
| JP2000021833A (en) * | 1998-07-02 | 2000-01-21 | Kaijo Corp | Spin cleaning device for semiconductor wafer |
| JP4377472B2 (en) * | 1999-03-08 | 2009-12-02 | 株式会社東芝 | Face image processing device |
-
2000
- 2000-12-27 CN CN00819023.2A patent/CN1261904C/en not_active Expired - Fee Related
- 2000-12-27 JP JP2002556837A patent/JP4476546B2/en not_active Expired - Fee Related
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101263074B (en) * | 2005-09-16 | 2013-03-27 | 奥蒂斯电梯公司 | Optically monitoring comb-line of escalators and moving walks |
| CN102292277A (en) * | 2008-12-11 | 2011-12-21 | 因温特奥股份公司 | Method for enabling the use of an elevator system by disabled persons |
| CN102292277B (en) * | 2008-12-11 | 2014-09-03 | 因温特奥股份公司 | Method for enabling the use of an elevator system by disabled persons |
| CN104743419A (en) * | 2013-12-25 | 2015-07-01 | 株式会社日立制作所 | Image monitoring device and elevator monitoring device |
| CN105989343A (en) * | 2014-09-25 | 2016-10-05 | 株式会社日立建筑系统 | Detection device and method of seeing-eye dog |
| CN111492371A (en) * | 2017-12-14 | 2020-08-04 | 三菱电机株式会社 | Retrieval system and monitoring system |
| CN111492371B (en) * | 2017-12-14 | 2023-05-26 | 三菱电机株式会社 | Search system and monitoring system |
| CN111507125A (en) * | 2019-01-30 | 2020-08-07 | 佳能株式会社 | Detection device and method, image processing device and system |
| CN113939469A (en) * | 2019-06-20 | 2022-01-14 | 三菱电机株式会社 | Control device for elevator |
| CN111362081A (en) * | 2020-04-15 | 2020-07-03 | 上海三菱电梯有限公司 | Elevator group management method and elevator group management system |
| CN111362081B (en) * | 2020-04-15 | 2022-06-17 | 上海三菱电梯有限公司 | Elevator group management method and elevator group management system |
Also Published As
| Publication number | Publication date |
|---|---|
| JPWO2002056251A1 (en) | 2004-05-20 |
| JP4476546B2 (en) | 2010-06-09 |
| CN1261904C (en) | 2006-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1605086A (en) | Image processing unit | |
| US7079669B2 (en) | Image processing device and elevator mounting it thereon | |
| US7627147B2 (en) | Method and apparatus for obtaining iris biometric information from a moving subject | |
| JPWO2020121973A1 (en) | Learning methods for object identification systems, arithmetic processing devices, automobiles, lamps for vehicles, and classifiers | |
| US9336439B2 (en) | System and method for the long range acquisition of iris images from stationary and mobile subjects | |
| US6690815B2 (en) | Image recognition apparatus and method | |
| US20030152251A1 (en) | Method and apparartus for picking up object being authenticated | |
| CN1647097A (en) | Method and system for classifying object in scene | |
| CN1498849A (en) | Camera for monitoring escalator and mobile footway | |
| JP2007504562A (en) | Method and apparatus for performing iris authentication from a single image | |
| CN1262239C (en) | Iris image pick-up and iris identifier | |
| CN1261904C (en) | Image processing device and elevator mounting it thereon | |
| CN1618079A (en) | System and method for iris identification using stereoscopic face recognition | |
| WO2005032887A3 (en) | Occupant detection system | |
| EP1533744A3 (en) | Apparatus and method for human distinction using infrared light | |
| JP2008113902A (en) | Apparatus and method for detecting eye opening degree | |
| JP2005100372A (en) | Imaging system and program | |
| JP2009237993A (en) | Image monitoring device | |
| JP2003150942A (en) | Eye position tracing method | |
| CN1689319A (en) | Photographing apparatus, photographing method and computer program | |
| JP2010169431A (en) | Passage detecting apparatus | |
| CN106994224A (en) | Move board | |
| JP4636314B2 (en) | Image processing apparatus and method, recording medium, and program | |
| JP2004226072A (en) | Shape measuring device | |
| KR20110058686A (en) | Method and device for detecting head position of traveling body |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060628 Termination date: 20141227 |
|
| EXPY | Termination of patent right or utility model |