JP2007272614A - Action analysis device using interlaced image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the following problem: though a demand that action of a minute living thing such as plankton, a fry, or a small fish is desired to be measured in a site wherein aquaculture is performed or a fishery-related research field, it is difficult to perform the action measurement because coping with a sharp change of motion of the minute living thing is impossible in conventional measurement technique. <P>SOLUTION: In this action analysis device, a characteristic image is measured as abnormal action from acquired interlaced images. Appearance numbers of observed normal action and the abnormal action, and fluctuation of the appearance numbers by a change of time are measured and recorded in an abnormal action appearance database 9. The appearance number 9a of the normal action and the appearance number 9b of the abnormal action are recorded in the abnormal action appearance database 9. By using the abnormal action appearance database 9, the appearance number, an appearance ratio, and the temporal fluctuation of the appearance number or the appearance ratio of the abnormal action are calculated, and they are set as an index of deciding health condition of the small fish, the fry, the plankton or the like. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a behavior analysis apparatus that can be used for behavior analysis of micro organisms such as larvae, fry, and plankton.

It is important to measure the movement of small organisms such as larvae, fry, and plankton in the field where aquaculture is conducted or in the fishery-related research field. Up to now, the movement of cultured larvae and larvae has been visually observed to make empirical judgments regarding the health of larvae and larvae, and the plankton's health status by observing plankton movements visually. Judgment was made. In addition, fisheries-related research institutions observe the movement of small organisms such as larvae, larvae, and plankton by visual observation, and record the movement of tiny fish such as larvae, larvae, and plankton on a grid of paper to analyze behavior. I was going. Alternatively, behaviors are measured with micro-organisms such as larvae, fry and plankton using a flow measurement device for measuring the flow of fluid and a general-purpose image processing device.
Kobayashi et al .: Development of a real-time digital image measurement system for two-dimensional flow fields, Transactions of the Japan Society of Mechanical Engineers (Part B), Vol. 55, No. 509 (1989)

  However, when humans make visual measurements, the measurement results vary greatly depending on human ability and fatigue, and quantitative measurements cannot be performed. In addition, it is difficult to carry out the method using grid-form paper because it requires long-term analysis and large-scale data analysis, and the measurement results may vary depending on the human ability and fatigue level of the analysis. It was difficult to measure. Furthermore, in analysis using a flow measurement device or general-purpose image processing device, it is necessary to analyze the behavior by individually tracking the target plankton and fish, but organisms such as larvae, fry and plankton are Since the movement is not uniform like the tracer particles used in measurement, there is a problem that when there is a sudden change in behavior peculiar to living things, the object is lost and measurement cannot be performed.

  In order to solve the above-described problems, the first invention is that the image data obtained from a TV camera, a DVD, a digital video, an image file, a high-definition image file, or the like is periodically or not used as an interlaced image including an even screen and an odd screen. An image capture unit that captures periodically and an interlaced image consisting of even and odd screens are processed simultaneously, and an abnormal behavior detection function that detects abnormal behavior based on the difference between captured even and odd rows Record the location and time information of micro-organisms such as larvae, fry, plankton in the data in memory or hard disk, and the number of appearances, appearance ratio, number of appearances of abnormal behavior detected by abnormal behavior detection function during a certain period of time Measure abnormal behavior of small organisms such as larvae, larvae, plankton, etc. Preparative a behavioral analysis apparatus using an interlaced image wherein.

  Further, the second invention assigns individual ID numbers to larvae, fry and plankton in the image data in the behavior analysis apparatus according to claim 1, and positions, times, and The presence or absence of abnormal behavior is measured, and if abnormal behavior is not detected, the direction and range in which the target organism is likely to move are determined from the image data acquired at regular or irregular time intervals. Determined from the moving direction and moving speed so far, the minute organisms after a predetermined time interval are related to the ID number from the image data, and if abnormal behavior is detected, the range in which the minute organisms will move is set This is a behavior analysis device using an interlaced image that can continuously track the behavior of a plurality of micro-organisms by changing the search.

Furthermore, the third invention is the behavior analysis apparatus using the interlaced image according to claim 2, wherein the position coordinates of one or more objects are recorded in a memory or a hard disk for each individual ID, and each recorded individual ID is recorded. This is a behavior analysis apparatus using an interlaced image characterized by functions of analyzing, sorting, and extracting behavior features of an object based on changes in time and position.

  In the present invention, in order to efficiently analyze the behavior of plankton, fish, etc., it is possible to detect signs of plankton or fish behavior change by simultaneously analyzing an interlaced image composed of even and odd images. Is related to the behavior analysis device. Focusing on the phenomenon that plankton, fish, etc. often change their behavior pattern, it is often accompanied by a sudden movement, and it is characterized by detecting the sudden movement as a sign of abnormal behavior. This phenomenon is characterized by the rapid movement of the object that can be used as a sign of changing the behavior pattern by displaying the interlaced image in which the even and odd screens are transmitted alternately, even and odd screens at the same time. Observed as a typical image. This image is observed as a jagged image of the periphery of the object because the position of the object measured as an even screen and an odd screen of the image is greatly shifted due to a sudden action. In the present invention, the abnormal behavior is detected by detecting the characteristic image generated in the video obtained by photographing the object that is the sign of the abnormal behavior by processing the even image and the odd image simultaneously. When a characteristic image is generated in an object image, it is likely that the object will behave differently from the previous action, so it is detected as abnormal behavior and the object that has been adopted so far is tracked By changing the search range necessary for tracking, tracking errors and tracking errors can be reduced. In general, after a characteristic image that is a sign of abnormal behavior is obtained, the traveling direction of the micro-organism changes greatly, or the speed changes greatly, so that the search range is expanded.

  One embodiment of the present invention is shown in FIG. In this embodiment, a micro-organism is put in a shallow liquid container such as a petri dish, and the measurement is performed by the behavior analysis apparatus using the interlaced image of the present invention.

Embodiments of the present invention will be described below with reference to the drawings. In the embodiment shown in FIG. 1, an analog TV camera 1a measures an image obtained by putting a micro organism in a shallow liquid container 3 such as a petri dish together with seawater or fresh water 2 and irradiating light with a light or a laser 4 from below. . The image signal of the analog TV camera is converted into a digital image signal by the analog image digital image converter 5 and sent to the behavior analysis device 6 using an interlaced image. In the behavior analysis apparatus 6 using the interless image, the behavior analysis of the target micro-organism is performed by processing the interlaced image including the even-numbered screen and the odd-numbered screen. The behavior analysis apparatus using the interlaced image includes a CPU 6a, an interface 6b, a hard disk 6c, a memory 6d, and a monitor 6e, and analyzes the behavior of the micro-organism using the interlaced image.

  FIG. 2 is an explanatory diagram of an acquired image according to the embodiment. An analysis method will be described in the case of using a general video in which an even screen and an odd screen are transmitted at a time interval of 1/60 sec. In this case, an interlaced image composed of even and odd screens is sent in 1/30 sec. The behavior analysis apparatus takes in even-numbered screens and odd-numbered screens that can be acquired every 1/30 sec and performs analysis. Using the example, when no abnormal behavior occurs, a video 7a as shown in FIG. 2 is observed. The micro organisms measured here can be measured as an image 8a by synthesizing the even screen and the odd screen.

  FIG. 3 is an explanatory diagram of an abnormal action image for detecting the embodiment. In the embodiment, the video 7b is observed when abnormal behavior occurs. In other words, because the movement of micro-organisms temporarily increased, the difference between the even-numbered screen and the odd-numbered screen becomes large and is observed as a characteristic image. According to the present invention, this characteristic image is extracted by observing an interlaced image composed of an even-numbered screen and an odd-numbered screen, and abnormal behavior of a micro-organism is detected.

FIG. 4 is an explanatory diagram regarding the appearance measurement of abnormal behavior in the embodiment. In the first aspect of the present invention, the number of appearances of normal behavior and abnormal behavior observed using the embodiment and the change in the number of appearances due to changes in time are measured and recorded in the abnormal behavior database 9. The abnormal behavior database 9 records the number of normal behaviors 9a and the number of abnormal behaviors 9b. By using this abnormal behavior appearance database 9, the number of appearances of abnormal behavior, the appearance ratio, the temporal variation of the number of appearances and the appearance ratio are calculated, and used as an index for judging the health status of larvae, fry, plankton, and the like.

  FIG. 5 is an explanatory diagram of a control computer block diagram and a database according to the embodiment. As image data, not only an image of an analog TV camera but also an image of a digital video camera 1b, a DVD, a digital video, an image file, or a high-definition image file can be used. According to the second aspect of the present invention, the video from the analog television camera 1 a is sent to the behavior analysis device 6 through the analog image digital image converter 5. Images from the digital video camera are sent directly to the behavior analysis device. In the behavior analysis device, individual ID numbers are assigned to larvae, fry, and plankton in the image data, and the position, time, and presence / absence of abnormal behavior are measured periodically or irregularly for each ID number of the countermeasure. Record in the association database 10. In this database, ID number 10a, measurement time 10b, horizontal coordinate 10c, vertical coordinate 10d, presence / absence of abnormal behavior 10e are recorded and used for associating micro-organisms recorded between images having different measurement times.

FIG. 6 is a flowchart of the embodiment. When measurement is started, an interlaced image is first captured. Next, the ID number, position, and measurement time of the object are recorded. A characteristic image is extracted from the interlaced image to determine whether the behavior is abnormal behavior or normal behavior. If the behavior is normal behavior, a corresponding object is detected and then a new interlaced image is captured. When it is determined from the characteristic video of the interlaced image that the target micro-organism is abnormal behavior, the search range is changed and the even screen and the odd screen are separated. This is to avoid a large error between the even-numbered screen and the odd-numbered screen when an abnormal action is detected, and processing simultaneously. After detecting the corresponding objects on the even and odd screens, a new interlaced image is captured.

FIG. 7 is an explanatory diagram for determining the search range of the embodiment. When the micro organism 8a performing normal behavior is observed, after predicting the traveling direction and the traveling speed from the past movement history, the next interlaced image is acquired, and the search is performed within the search range of 11a. Corresponding to the image 12a of the organism. On the other hand, since the micro organism 8b in which the abnormal behavior is detected is likely to perform a motion that is significantly different from the previous motion in the next acquired image, the search range 11b centered on the current position is set. Then, the image is associated with the image 12b of the microbiological tsu. By detecting the abnormal behavior using the interlaced image, it is possible to associate the micro-organism 12b having a behavior that is significantly different from the conventional behavior.

FIG. 8 is an explanatory diagram of a database used for action classification in the embodiment. According to the third aspect of the present invention, the movement coordinates of one or a plurality of objects are rearranged along the change in measurement time for each individual ID and recorded as the behavior analysis database 13 in a memory or a hard disk. Using the individual ID 13a, the measurement time 13b, the horizontal coordinate 13c at the measurement time, and the vertical coordinate 13d at the measurement time, a position table along the time for each individual ID is obtained, and the movement speed change of the target micro-organism It is possible to acquire changes in the direction of travel.

"Effect of the embodiment"
According to this embodiment, abnormal behaviors of a plurality of micro-organisms can be observed by a simple process using an interlaced image. In addition, since the movement of a large number of micro-organisms can be analyzed individually and the movement of the micro-organisms for each individual ID can be analyzed, it can be used for behavior analysis of a large number of micro-organisms that are problematic in fishery-related research institutions. Furthermore, by analyzing and classifying the behavior, it can be used to identify larvae, fry, plankton, etc. that have similar movements but different movements.

"Other embodiments"
In another embodiment shown in FIG. 9, the television camera 1a is waterproofed and can be used in a large water tank by irradiating the light or the laser 4 from the side.

  In another embodiment shown in FIG. 10, only the video is acquired at the site by the digital video camera 1 b and later analyzed by the behavior analysis device 6. Since only the camera needs to be brought into the field, the complexity of the measurement work can be avoided. Also, a DVD or a high-definition disc (Blu-ray disc, HD disc) may be used.

By using the present invention, it becomes easy to analyze the behavior of larvae, fry and plankton in the field where aquaculture is carried out or in a fishery-related research field. Since the behavior of larvae and larvae can be used as an index representing the health status of larvae and larvae, it can contribute to the advancement of farmed fish. Moreover, it can contribute to the stable supply of the feed for cultured fish by performing the behavioral analysis of the plankton used as a food of cultured fish by the Example of this invention. In addition, by using the embodiments of the present invention, it is possible to improve the efficiency and speed of fishery-related academic research.

It is a block diagram of an Example. It is explanatory drawing of the acquired image of an Example. It is explanatory drawing of the abnormal action image which performs the detection of an Example. It is explanatory drawing of appearance measurement of abnormal action in an Example. It is explanatory drawing of the control computer block diagram and database of an Example. It is a flowchart of an Example. It is explanatory drawing of the search range determination of an Example. It is explanatory drawing of the database used for the classification | category of the action of an Example. It is a block diagram of the other Example using a waterproof camera. It is a block diagram of the other Example using a digital video.

Explanation of symbols

1a Analog TV camera
1b Digital video camera 2 Seawater or fresh water in a container
3 Liquid container 4 Light or laser
5 Analog image digital image converter 6 Behavior analysis device 6a CPU
6b interface 6c hard disk 6d memory 6e monitor 7a video in which abnormal behavior is not observed 7b video in which abnormal behavior is observed 8a video of micro organisms in normal behavior 8b video of micro organisms in abnormal behavior 9 abnormal behavior database 9a normal behavior 9b Number of occurrences of abnormal behavior 10 Behavior association database 10a Behavior association database ID number 10b Action association database measurement time 10c Behavior association database horizontal coordinate 10d Behavior association database vertical coordinate 10e Presence of behavior association database abnormal behavior 11a Normal behavior Search range 11b Search range in abnormal behavior 12a Corresponding video in normal behavior 12b Corresponding video in abnormal behavior 13 Behavior analysis database 13a Behavior analysis database ID number 13b Motion analysis database measurement time 13c Behavior analysis database horizontal coordinate 13d Behavior analysis database vertical coordinate n Arbitrary ID number Tn, m Time measured at the mth ID number n Xn, m Horizontal coordinate measured at the mth ID number n Yn , M ID coordinate n measured vertical coordinate

Claims (3)

An image capture unit that captures image data obtained from a TV camera, DVD, digital video, image file, high-definition image file, etc. regularly or irregularly as an interlaced image consisting of even and odd screens, and from even and odd screens It is equipped with an abnormal behavior detection function that detects abnormal behavior based on the difference between even and odd rows that are captured simultaneously, and the location of micro-organisms such as larvae, fry, plankton, etc. in the image data By recording the time information in the memory or hard disk and calculating the number of appearances of abnormal behavior detected by the abnormal behavior detection function during a certain period of time, the appearance ratio, the temporal variation of the number of appearances and the appearance ratio, Lines using interlaced images characterized by measuring abnormal behavior of small organisms such as fry and plankton Analysis apparatus. The behavior analysis apparatus using the interlaced image according to claim 1, wherein individual ID numbers are assigned to larvae, fry and plankton in the image data, and the position and time for each ID number of the target object are regularly or irregularly assigned. If abnormal behavior is not detected, the direction and range in which the target micro-organism will be moving from the image data acquired at regular or irregular time intervals are determined. Predicted from the moving direction and moving speed of the past, the micro organism after a predetermined time interval is associated with the ID number from the image data, and if abnormal behavior is detected, the micro organism will be moving By changing the range setting and performing searches, it is possible to respond to sudden changes in behavior unique to micro-organisms, and it is possible to continuously track the behavior of multiple micro-organisms Behavioral analysis device using the image. The behavior analysis apparatus using an interlaced image according to claim 2, wherein the measurement time and position coordinates of one or more objects are recorded in a memory or a hard disk for each individual ID, and the time and position for each recorded individual ID. A behavior analysis apparatus using an interlaced image characterized by the function of analyzing, sorting, and extracting features of an object based on changes in coordinates.

Images