Improved fish behavior analysis tank based on camera shooting technology and behavior parameter evaluation method

Abstract

The invention belongs to the field of biological and ecological toxicity monitoring and aquatic organism behavior analysis, and particularly discloses an improved fish behavior analysis tank based on a camera shooting technology and a behavior parameter evaluation method. The problems of monitoring dead angles, false positive and the like caused by image deformation when a common camera is used for monitoring and identifying a fish body are solved. The device comprises a cylinder body and a partition plate assembly, the inner wall of the partition plate assembly is parallel to a camera shooting light path, the partition plate assembly is composed of a plurality of partition plates with gradually-changed thicknesses, and the thicknesses of the partition plates are gradually reduced in the camera shooting light path direction. The partition plate assembly comprises a left partition plate and a right partition plate, excrement grooves are formed in the bottoms of the left partition plate and the right partition plate, and a plurality of evenly-distributed water inlet and outlet through holes are formed in the upper middle portions of the left partition plate and the right partition plate. When the camera shooting light path is in the vertical direction, the partition plate assembly further comprises a front side partition plate and a rear side partition plate. When the camera shooting light path is in the horizontal direction, the partition plate assembly further comprises a bottom partition plate and a top partition plate.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of biological and ecological toxicity monitoring and aquatic organism behavior analysis, and particularly relates to a fish behavior analysis tank improved based on camera technology and a behavior parameter evaluation method. BACKGROUND

[0002] With the progress of society and industrial development, more and more chemicals are developed to improve the production and living standards of human beings. This also makes more and more chemicals enter the environmental water bodies in various ways (discharge, leakage, use, etc.). These exogenous chemicals will have multi-aspect (neurological, endocrine, reproductive, etc.), multi-level (community, individual, organ, tissue, cell, biological molecule, etc.) and multi-level (biological enrichment, biological amplification, biological accumulation, etc.) effects on aquatic organisms, threatening the ecological health of aquatic organisms and further adversely affecting human health.

[0003] In order to evaluate the water body environment and water body ecological quality, water quality monitoring technology emerges as the times require, and different water quality parameters (total nitrogen, total phosphorus, ammonia nitrogen, BOD, COD, pH, etc.) limits of different water bodies have been stipulated by countries all over the world. However, with the rapid development of chemical technology, the conventional water quality monitoring indicators cannot accurately reflect the harmfulness of water bodies to the environment, and some pollutants will produce greater toxic effects at very low concentration levels without causing large fluctuations in monitoring parameters.

[0004] The whole effluent toxicity (WET) test is a water quality monitoring method proposed by the EPA, which can obtain toxicity data of wastewater to aquatic organisms without identifying specific pollutants, and is a direct and efficient method for determining the harm of wastewater to aquatic organisms. Similarly, this method can also be used for water toxicity evaluation and toxicity online monitoring.

[0005] Toxicity online monitoring technology is a monitoring technology based on the stress response of model organisms to abnormal water quality. By monitoring the changes in the behavior of model organisms, water pollution early warning can be achieved. Common model organisms include luminescent bacteria, algae, fish, and shellfish. Among them, luminescent bacteria are the most widely used and can be used to evaluate the acute toxicity of water. The Technical Requirements for Biological Toxicity Water Quality Automatic Online Monitoring Instrument Luminescent Bacteria Method DB44 / T 1946-2016 specifies the technical requirements for using luminescent bacteria to monitor the biological toxicity of water sources such as surface water and groundwater online. However, due to its low trophic level, it is difficult to directly reflect the comprehensive impact of water quality changes on the entire aquatic ecosystem, and other types of aquatic organisms need to be used for toxicity monitoring to supplement the data. Fish, as a species with a higher trophic level in the aquatic ecosystem, can adapt to slightly polluted water. Unlike algae, daphnia, and microorganisms, the toxicity of water to fish can be determined by behavior without reaching a lethal effect, truly achieving environmental friendliness in the monitoring method. Small fish model organisms such as zebrafish, medaka, and rare gobi- carp are small in size, have short development and experimental cycles, are low-cost, and require small amounts of experimental drugs, making them widely used in toxicity research. For example, Chinese invention patent CN105259325A discloses an online monitoring and early warning device using zebrafish as a model organism.

[0006] There are various techniques for using fish as model organisms for biological toxicity online monitoring, including camera tracking monitoring technology, four-stage impedance monitoring technology, respiratory determination technology, bioelectric effect determination technology, electrocardiogram and electroencephalogram determination technology, and combined technology with other techniques. Among these monitoring technologies, camera tracking monitoring technology is more intuitive, has low basic requirements for hardware equipment, can monitor a large number of items, and is an important medium for coupling other technologies. Infrared camera tracking is a method that uses an infrared camera to capture infrared light and forms an image based on the obtained light intensity. It has been applied to behavior analysis of juvenile fish and adult fish. Water and its container are determined to be bright due to high light transmittance, while the fish body and the container connection part are determined to be dark due to low light transmittance. By properly framing the analysis area, fish can be identified.

[0007] Common cameras include ordinary cameras and wide-angle cameras. Wide-angle cameras can affect the identification and quantitative analysis of fish bodies due to image edge distortion. Ordinary cameras have a wide field of view that expands with distance, but are easily affected by water flow fluctuations and reflections. Infrared cameras, on the other hand, respond to the connection between the container and the water surface, affecting the identification of fish bodies, resulting in the loss of the monitoring range of the upper water body. Pollutants can cause fish to escape and float upwards, entering the monitoring blind area and causing false positives or invalid data. However, existing technologies have no low-cost solutions to these problems. SUMMARY

[0008] An object of the present application is to provide an improved fish behavior analysis tank based on camera technology, which effectively solves the problem of monitoring dead angles and false positives caused by image distortion when using the commonly used camera to monitor and identify fish.

[0009] To solve the above technical problems, the technical solution adopted by the present application is:

[0010] An improved fish behavior analysis tank based on camera technology, comprising a tank body and a partition assembly, the partition assembly is arranged in the tank body, the inner wall of the partition assembly is designed in parallel to the camera light path, the partition assembly is composed of multiple partitions with gradually changing thickness, the internal space of the partition assembly is an experimental area, and the thickness of the partition gradually decreases along the direction of the camera light path.

[0011] The partition assembly includes a left partition and a right partition, the inner surfaces of the left partition and the right partition are both trapezoidal in cross-section along the direction of the water flow, the bottom surfaces and side edges of the left partition and the right partition are adapted to and fit the bottom wall of the tank body, the bottom of the left partition and the right partition is provided with a feces groove, and the middle and upper parts of the left partition and the right partition are both provided with multiple uniformly distributed water inlet and outlet through holes.

[0012] When the camera light path is in the vertical direction, the partition assembly further includes a front partition and a rear partition, and the front partition and the rear partition are tightly fitted with the front side wall and the rear side wall of the tank body, respectively.

[0013] When the camera light path is in the horizontal direction, the partition assembly further includes a bottom partition and a top partition.

[0014] Further, the thickness calculation formula of the partition is: In the formula, a represents the thickness difference between the camera head end and the camera head end of the partition, h1 represents the distance between the camera head and the tank wall away from the camera head end, h2 represents the height of the partition in the camera light path direction, and l represents the distance between the two partitions away from the camera head.

[0015] Further, when the camera light path is in the vertical direction, a top cover plate is arranged on the partition assembly to prevent fish from escaping the experimental area.

[0016] Further, when the camera light path is in the horizontal direction, the bottom partition is the bottom wall of the tank body.

[0017] Further, the water inlet area and the water outlet area are included, the water outlet area is provided with an internal circulation filter, and the water inlet area is provided with an aeration device and a sewage dispersion system.

[0018] Further, the sewage dispersion system comprises a pollutant inlet pipe and a circulating water pipe and an aeration assembly arranged vertically downward, the pollutant inlet pipe and the circulating water pipe are closely fitted, and the outlet of the pollutant inlet pipe is higher than the outlet of the circulating water pipe.

[0019] Further, the inner surface of the partition assembly is in the shape of a prism.

[0020] Further, the cylinder is made of transparent glass or transparent plastic material, and the partition and the top cover plate are made of transparent plastic material.

[0021] Further, the plastic is endocrine disruptor additive-free acrylic plastic.

[0022] Further, the volume of the cylinder is not less than 10 L, and the length, width and height of the cylinder are all not more than 50 cm.

[0023] Another object of the present application is to provide a behavior parameter evaluation method using the fish behavior analysis cylinder based on the improved camera technology.

[0024] (I) Realized by adding an auxiliary camera perpendicular to the analysis camera through a 2D behavior analysis system.

[0025] I. The analysis camera is vertically arranged, and the auxiliary camera is horizontally arranged.

[0026] The fish behavior analysis cylinder is used to quantitatively evaluate the behavior parameters when the camera light path is in the vertical direction, including the following steps:

[0027] A1. Fix the vertical analysis camera above the cylinder and at the center vertical line of the experimental area.

[0028] A2. Fix the horizontal auxiliary camera in front of the cylinder and at the center horizontal line of the experimental area.

[0029] A3. Record the fish behavior video in the vertical and horizontal directions, and perform trajectory recognition and behavior parameter output. The actual movement distance of the fish in the horizontal direction between the i-th frame and the i+1-th frame is x i : In the formula, a represents the thickness difference between the top end and the bottom end of the left partition or the right partition, h1 represents the distance between the analysis camera and the bottom wall of the cylinder, h2 represents the height of the partition in the direction of the camera light path of the analysis camera, l represents the distance between the bottom end of the left partition and the bottom end of the right partition, and S iS represents the distance of the fish body output on the video image from the i-th frame to the i+1-th frame, that is, the length of the line segment formed by the projection of the starting point and the ending point of the actual motion trajectory on the bottom wall of the cylinder, that is, the actual horizontal motion distance of the fish, H i value, x i S represents the distance of the fish body output on the video image from the i-th frame to the i+1-th frame, that is, the length of the line segment formed by the projection of the starting point and the ending point of the actual motion trajectory on the bottom wall of the cylinder, that is, the actual horizontal motion distance of the fish, H i and H i+1 respectively represent the height of the fish body in the i-th frame and the i+1-th frame output by the auxiliary camera.

[0030] If the frame rate of the camera is 30 frames / s, then the motion distance of the fish in the horizontal direction per second is: wherein, X j is the total motion distance of the fish in the horizontal direction per second.

[0031] II, the analysis camera is horizontally shot, and the auxiliary camera is vertically shot.

[0032] The fish behavior analysis tank is used to quantitatively evaluate the behavior parameters when the camera light path is in the horizontal direction. The analysis target is the behavior parameters when the fish body floats or sinks. The behavior parameters include but are not limited to floating or sinking speed, motion distance in the vertical direction, residence time in the upper half area, residence time in the lower half area, and swimming height. The following steps are included:

[0033] B1, the horizontal analysis camera is fixed in front of the cylinder body, and the shooting center of the analysis camera is located on the center horizontal line of the experimental area.

[0034] B2, the vertical auxiliary camera is fixed directly above the cylinder body, and the shooting center of the auxiliary camera is located on the center vertical line of the experimental area.

[0035] B3, record the fish behavior video in the vertical and horizontal shooting directions and perform trajectory recognition and behavior parameter output.

[0036] The vertical downward auxiliary camera records the vertical distance w of the i-th fish body from the cylinder wall close to the analysis camera i , the horizontal analysis camera records the height h of the i-th fish body from the bottom baffle in the image i , and the actual swimming height h i ' of the i-th fish body from the bottom baffle is calculated according to formula (4): wherein, W is the distance between the analysis camera and the cylinder wall away from the analysis camera, and w is the distance difference between the cylinder side wall close to the analysis camera and the cylinder side wall away from the analysis camera, that is, the distance between the front side wall of the cylinder and the rear side wall of the cylinder.

[0037] The floating or sinking distance between the i-th frame and the i+1-th frame: Ah i = h i+1 ' - h i '(5), wherein h i ' and h i+1 ' represent the actual swimming height of the fish from the bottom partition in the i-th frame and the i+1-th frame, respectively.

[0038] The average movement speed V and the movement rate v from the j-th second to the k-th second are:

[0039]

[0040] In the formula, V represents the average movement speed from the j-th second to the k-th second, and if the average movement speed is positive, it indicates that the fish has an upward swimming trend from the j-th second to the k-th second, and vice versa.

[0041] v represents the movement rate from the j-th second to the k-th second, and indicates the activity of the fish in the vertical direction, and the greater v is, the more active the fish is in the vertical direction, and vice versa.

[0042] Δt j-k represents the time interval, and n represents the frame rate of the camera used.

[0043] (II) is realized by a 3D behavior analysis system.

[0044] The fish behavior analysis tank simultaneously meets the requirements of a fish behavior analysis tank when the camera light path is horizontal and the requirements of a fish behavior analysis tank when the camera light path is vertical, and the fish behavior analysis tank does not need to be provided with a top partition with varying thickness, and the left and right partitions are gradually thinned in the vertical camera light path direction.

[0045] A first analysis camera and a second analysis camera are respectively arranged in the vertical and horizontal directions, the coordinates of each frame of 3D space are corrected according to formulas (8)-(13), and the trajectory is drawn and the 3D space distribution is analyzed according to the corrected 3D coordinates of each frame of image.

[0046] The actual movement distance of the fish in the horizontal direction between the i-th frame and the i+1-th frame is x i : In the formula, a represents the thickness difference between the top and bottom of the left or right partition, h1 represents the distance between the first analysis camera and the bottom wall of the tank, j2 represents the height of the partition in the camera light path direction of the first analysis camera, l represents the distance between the bottom of the left partition and the bottom of the right partition, and S iThe motion distance of the fish body output on the video image from the i-th frame to the i+1-th frame, that is, the horizontal line connecting the first camera and the i-th frame and the i+1-th frame respectively and extending to the horizontal plane of the cylinder bottom, and the line segment formed by the two intersection points of the horizontal plane of the cylinder bottom is S i The value of x i The length of the projection line of the starting point and the ending point of the actual motion trajectory on the bottom partition, that is, the actual horizontal motion distance of the fish, H i and H i+1 are the heights of the fish body in the i-th frame and the i+1-th frame output by the second analysis camera respectively.

[0047] If the frame rate of the first analysis camera is 30 frames / s, then the motion distance of the fish in the horizontal direction per second is: In the formula, X j is the total motion distance of the fish in the horizontal direction per second.

[0048] The actual swimming height h i of the fish body in the i-th frame from the bottom partition In the formula, W represents the distance between the second analysis camera and the cylinder wall away from the second analysis camera, and w represents the distance difference between the cylinder side wall close to the second analysis camera and the cylinder side wall away from the second analysis camera, that is, the distance between the front side wall of the cylinder and the rear side wall of the cylinder, and w i represents the vertical line segment distance of the fish body in the i-th frame recorded by the first analysis camera from the cylinder wall close to the second analysis camera.

[0049] Then the floating or sinking distance between the i-th frame and the i+1-th frame is Δh i = h i+1 '- h i '(11), in which h i ' and h i+1 ' represent the actual swimming height of the fish body in the i-th frame and the i+1-th frame from the bottom partition respectively.

[0050] The average motion speed V and the motion rate v from the j-th second to the k-th second are:

[0051]

[0052] In the formula, V represents the average motion speed from the j-th second to the k-th second, and if the average motion speed is positive, it indicates that the fish has an upward swimming trend from the j-th second to the k-th second, and vice versa.

[0053] v represents the motion rate from the j-th second to the k-th second, and represents the activity of the fish in the vertical direction, and the greater v is, the more active the fish is in the vertical direction, and vice versa.

[0054] αtj-k represents the time interval, n represents the frame rate of the camera used.

[0055] Compared with the prior art, the application has the beneficial technical effects that:

[0056] (1) The application ensures the uniform diffusion of pollutants while eliminating the dead angle of the monitoring range by designing a baffle assembly parallel to the camera light path, thereby achieving comprehensive coverage of behavior monitoring. Meanwhile, the proposed behavior parameter quantitative evaluation method can improve the accuracy of data, effectively reduce the error of fish movement parameters on different cross-sections caused by image deformation, and greatly eliminate false negative or false positive results caused by the camera light path, thereby providing technical support for reliable water quality monitoring in different water bodies, non-destructive toxicity evaluation of chemicals, and other fields.

[0057] (2) The application uses a thickness-graduated acrylic punched baffle as the water inlet and outlet baffle and the cylinder wall baffle, which can be coupled with a video behavior analysis system to locate the actual position according to the real-time video image, so that the activity range of fish is within the monitoring area. At the same time, it is beneficial to achieve the dispersion function of pollutants entering the water and provide sufficient activity space for fish, and it is difficult to analyze and observe behaviors such as escape and corner turning in small containers.

[0058] (3) The application avoids monitoring blind spots by setting a baffle assembly parallel to the monitoring camera light path, which is more cost-effective and simpler in hardware setup, and conforms to the lightweight principle.

[0059] With the baseline of the stable long-term monitoring behavior state of fish in aquaculture water, the difference in behavior state caused by the stress movement of fish due to abnormal water quality exposure reflects the real-time state of water quality change, which is used for monitoring the biological toxicity of abnormal water bodies.

[0060] With the baseline of the behavior state of fish in the target natural water body or long-term stable water discharged in line with national discharge standards, the ecological toxicity of abnormal water bodies can be reflected through abnormal behavior. Setting an appropriate baseline can avoid false positive results, while avoiding test errors caused by individual differences and reducing the use of animals, which is more in line with the 3R principle.

[0061] (4) The application has good application prospects in fish behavior analysis technology based on camera tracking technology. As a fixed component of the fish behavior analysis system based on camera tracking, it can effectively solve the problem of monitoring dead angles caused by the inherent technical characteristics of the camera. Combined with the behavior parameter evaluation method proposed in the application, false positives and invalid results can be greatly avoided, especially in the field of toxicology of pollutants, chemicals, and drugs. At the same time, the application has application value in the field of online monitoring of biological toxicity and ecological toxicity of polluted water, especially in the petrochemical industry, which can reduce the false alarm rate. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 is a structural schematic diagram of a fish behavior analysis tank with a vertical camera light path according to the present application.

[0063] Figure 2 is Figure 1 is a structural schematic diagram of all the partitions in the present application.

[0064] Figure 3 is a structural schematic diagram of a fish behavior analysis tank with a horizontal camera light path according to the present application and its left partition, right partition and top partition.

[0065] Figure 4 is a structural schematic diagram of a fish behavior analysis tank with a 3D behavior analysis system for fish behavior parameter evaluation according to the present application and its partition assembly.

[0066] Figure 5 is Figure 4 is a left view of the left partition in the present application.

[0067] Figure 6 is a pure water inflow zebra fish movement time-inflow time schematic diagram of Example 1 in the present application.

[0068] Figure 7 is a pure water inflow zebra fish movement time-inflow time schematic diagram of Example 1 in the present application.

[0069] List of reference signs: 1, tank body; 4, left partition; 5, right partition; 6, feces groove; 7, water inlet and outlet through hole; 8, front partition; 9, rear partition; 10, bottom partition; 11, top partition. DETAILED DESCRIPTION

[0070] Embodiment 1

[0071] A fish behavior analysis tank improved based on camera technology, comprising a tank body 1 and a partition assembly, the partition assembly is arranged in the tank body 1, preferably, the inner surface of the partition assembly is in the shape of a truncated pyramid as a whole.

[0072] The inner wall of the partition assembly is designed to be parallel to the camera light path, the partition assembly is composed of multiple partitions with gradually changing thickness, the internal space of the partition assembly is an experimental area, and the thickness of the partition gradually decreases along the direction of the camera light path.

[0073] The partition assembly comprises a left partition 4 and a right partition 5, the inner surfaces of the left partition (water inlet partition) 4 and the right partition (water outlet partition) 5 are both trapezoidal in cross section along the water flow direction, and the bottom surfaces and side edges of the left partition 4 and the right partition 5 are adapted to the bottom wall of the cylinder body. The bottom of the left partition 4 and the right partition 5 is provided with a fecal groove 6, which is used to quickly discharge normal excretion of fish by using the hydraulic effect and the gravity effect. Since the upper part of the partition is thick and the lower part is thin, the contaminants are more likely to flow from the lower part than from the upper part. In order to make the water inlet more homogeneous, the middle and upper parts of the left partition 4 and the right partition 5 are provided with a plurality of evenly distributed water inlet and outlet through holes 7.

[0074] As shown in Figure 1 and Figure 2 When the camera light path is in the vertical direction, the partition assembly further comprises a front partition 8 and a rear partition 9, which are tightly fitted with the front wall and the rear wall of the cylinder body 1, respectively.

[0075] Preferably, when the camera light path is in the vertical direction, the partition assembly is provided with a top cover plate for preventing fish from escaping from the experimental area.

[0076] As shown in Figure 3 When the camera light path is in the horizontal direction, the partition assembly further comprises a bottom partition 10 and a top partition 11.

[0077] Preferably, when the camera light path is in the horizontal direction, the bottom partition 10 is the bottom wall of the cylinder body.

[0078] In some specific embodiments, the cylinder body is made of high-transparency glass or high-transparency plastic material, and the partitions and the top cover plate are both made of high-transparency plastic material.

[0079] Preferably, the plastic is acrylic plastic without the addition of endocrine disruptors such as bisphenol A.

[0080] In some specific embodiments, the volume of the cylinder body is not less than 10L, and the length and the width of the cylinder body are both not more than 50cm. In order to reduce the error caused by the deformation of the camera shooting and the partition, the height of the cylinder body is not more than 50cm, that is, the length of the cylinder body in any shooting direction (vertical and horizontal directions) is not more than 50cm.

[0081] The thickness of the partition is determined according to the actual size of the cylinder body and the height of the camera, and the specific calculation formula is: In the formula, a represents the difference in thickness (cm) between the end of the baffle closest to the camera and the end farthest from the camera, h1 represents the distance (cm) from the camera to the cylinder wall farthest from the camera, h2 represents the height (cm) of the baffle in the direction of the camera light path, and l represents the distance (cm) between the two baffle ends farthest from the camera, i.e., the distance between the baffle ends of the maximum fish activity area. It should be noted that the variables herein are all parameters in the same view, such as the thickness of the left and right baffles in the front view. Specifically, as shown in FIG. 2, Figure 1 Figure 1 The left baffle (water inlet baffle) 4 and the right baffle (water outlet baffle) 5 shown in FIG. 2 are a pair of opposing baffles, the distance between the parallel lines at the bottom of the baffles (inside the tank bottom) is l, and the difference between the top thickness and the bottom thickness of the left and right baffles is a, Figure 1 In FIG. 2, the bottom thickness of the baffle is 0, so the top thickness is a, h1 is the vertical distance from the camera to the tank bottom, and h2 is the height of the baffle (the height of the tank).

[0082] In some specific embodiments, considering that the water body should be suitable for fish to live for a long time, an internal circulation filter is provided in the water outlet area, and an aeration device and a sewage dispersing system are provided in the water inlet area. The homogeneity of the water body and the removal of excrement are accelerated by the hydraulic flow accompanying the operation of the internal circulation filter and the aeration device.

[0083] The sewage dispersing system includes a contaminant inlet pipe and a vertically downward circulating water pipe and an aeration assembly, the contaminant inlet pipe and the circulating water pipe are closely fitted, and the outlet of the contaminant inlet pipe is higher than the outlet of the circulating water pipe. After the contaminant is introduced, it is mixed uniformly downward under the circulating power of about 1000 L / h of the circulating water, and irregular bubbles generated by aeration can also carry sewage upward, generating a two-way dispersing power, which can achieve pre-mixing in the water inlet area.

[0084] Taking the vertical direction of the camera light path as an example, due to the wide upper part and narrow lower part of the baffle, the resistance of the upper part is larger. In order to prevent the contaminant from being concentrated in the lower part due to the large resistance, a plurality of uniformly distributed water inlet and outlet holes are provided in the upper middle part of the left baffle (water inlet baffle) and the right baffle (water outlet baffle), the hole diameter is 2 mm, and the hole spacing is 2 mm. The diluted contaminant enters the experimental area through the water inlet and outlet holes, and spreads in the horizontal and vertical directions under the action of gravity and the circulating water pump in the water outlet area. Since there are fish swimming in the experimental area, the uncertainty of the contaminant is enhanced, and the dispersing process of the contaminant is accelerated.

[0085] Embodiment 2

[0086] ​The fish behavior analysis tank based on the improved camera technology in embodiment 1 is fully adapted to 2D camera tracking behavior analysis by the upper camera, and can also be coupled with the lateral camera for 3D behavior analysis. It is worth noting that the upper part of the lateral camera is not corrected for camera light path, so there is still a part of the monitoring blind area, but the overall analysis area of the monitoring dead angle has been greatly reduced.

[0087] Similarly, the baffle assembly shown in FIG. 1 can be used to correct the horizontal shooting light path for analyzing the movement rules of fish in the vertical direction, and the size calculation method is also applicable to formula (1). Such a baffle assembly is adapted for 2D camera tracking behavior analysis by the camera shooting in the horizontal direction. Figure 3

[0088] A behavior parameter evaluation method using the fish behavior analysis tank based on the improved camera technology in embodiment 1 is needed to perform appropriate size conversion to obtain accurate behavior parameter data, because the size of each cross section on the shooting line is not consistent, and the shooting monitoring selected area is fixed.

[0089] The behavior parameter evaluation method can be assisted by adding an auxiliary camera perpendicular to the analysis camera to the 2D behavior analysis system, or it can be realized by a 3D behavior analysis system. The specific method is as follows:

[0090] (1) It is realized by adding an auxiliary camera perpendicular to the analysis camera to the 2D behavior analysis system.

[0091] (I) The analysis camera shoots in the vertical direction, and the auxiliary camera shoots in the horizontal direction.

[0092] The fish behavior analysis tank based on the improved camera technology in embodiment 1 is fully adapted to 2D camera tracking behavior analysis by the upper camera, and can also be coupled with the lateral camera for 3D behavior analysis. It is worth noting that the upper part of the lateral camera is not corrected for camera light path, so there is still a part of the monitoring blind area, but the overall analysis area of the monitoring dead angle has been greatly reduced. Figure 1 Figure 2 The fish behavior analysis tank based on the improved camera technology in embodiment 1 is fully adapted to 2D camera tracking behavior analysis by the upper camera, and can also be coupled with the lateral camera for 3D behavior analysis. It is worth noting that the upper part of the lateral camera is not corrected for camera light path, so there is still a part of the monitoring blind area, but the overall analysis area of the monitoring dead angle has been greatly reduced.

[0093] A1, fix the vertical analysis camera directly above the tank body, and the shooting center of the analysis camera is located on the central vertical line of the experimental area.

[0094] The height of the analysis camera from the tank bottom is h1, the height of the baffle is h2, the distance difference between the top and bottom of the baffle is a, the bottom distance of the maximum activity area (experimental area) of the fish is l, and the top distance of the maximum activity area (experimental area) of the fish is l-2a. The height of the tank body, the analysis camera and the size of the baffle all need to meet the relationship described in formula (1).

[0095] ​​A2, fix the horizontal auxiliary camera in front of the cylinder body, and the shooting center of the auxiliary camera is located on the center horizontal line of the experimental area, so that the video image of the experimental area in the horizontal direction is located at the center of the overall shooting area. Since the horizontal camera is an auxiliary camera, the distance is not strictly required, and only the position of the fish body can be identified.

[0096] A3, record the fish behavior video in the vertical and horizontal shooting directions, and perform trajectory recognition and behavior parameter output.

[0097] The video output fish trajectory and the motion distance according to the set scale, the scale is set to the distance from the water baffle bottom, that is, the side length of the experimental area in the video image is set to l, and the output interval is every frame. The motion distance of the fish body in the video image between the i-th frame and the i+1-th frame is S i . The auxiliary video outputs the height H i of the fish body from the tank bottom in the i-th frame image.

[0098] The shooting baffle will have a slight arc deformation, but the deformation can be ignored under the condition that the shooting distance is sufficient, so the shooting area is a nearly conical area, and the activity of the fish in all activity directions can be calculated by the following formula.

[0099] The actual motion distance of the fish in the horizontal direction between the i-th frame and the i+1-th frame is x i : In the formula, a represents the thickness difference between the top and bottom of the left or right baffle, h1 represents the distance between the analysis camera and the tank bottom wall, h2 represents the height of the baffle in the direction of the analysis camera, l represents the distance between the bottom of the left baffle and the bottom of the right baffle, S i represents the motion distance of the fish in the video image from the i-th frame to the i+1-th frame, that is, the analysis camera and the i-th frame and i+1-th frame are connected and extended to the bottom wall of the tank, and the two intersection points with the bottom wall of the tank form a segment S i value, x i represents the length of the projection line of the starting point and the ending point of the actual motion trajectory on the bottom wall of the tank, that is, the actual horizontal motion distance of the fish, H i and H i+1 are the heights of the fish in the i-th frame and the i+1-th frame output by the auxiliary camera, respectively.

[0100] If the frame rate of the camera is 30 frames / s, then the motion distance of the fish in the horizontal direction per second is: In the formula, X jLet be the total horizontal distance traveled by the fish in the j-th second. By accumulating the distance traveled per second, the distance traveled over longer time intervals can be obtained, and the speed of movement within the corresponding time interval can be calculated.

[0101] (II) Analyze the camera shooting in the horizontal direction and the auxiliary camera shooting in the vertical direction.

[0102] When the camera's optical path is horizontal, such as Figure 3 The fish behavior analysis tank shown quantitatively assesses behavioral parameters. The analysis target is the behavioral parameters of the fish when it rises or sinks. These behavioral parameters include, but are not limited to, rising or sinking speed, vertical distance, time spent in the upper region, time spent in the lower region, and swimming height. The analysis includes the following steps:

[0103] B1. Fix the horizontal analysis camera directly in front of the cylinder and ensure that the camera's shooting center is located on the horizontal center line of the experimental area.

[0104] B2. Fix the vertical auxiliary camera directly above the cylinder body, with the camera's shooting center located on the vertical line of the center of the experimental area.

[0105] B3. Record fish behavior videos from both vertical and horizontal shooting directions, and perform trajectory recognition and output behavior parameters.

[0106] A vertically downward auxiliary camera records the distance w of the fish's body from the vertical segment of the tank wall closest to the analysis camera in the i-th frame. i The horizontal analysis camera records the height h of the fish's body from the bottom partition in the i-th frame of the image. i The actual swimming height h of the fish in the i-th frame from the bottom partition is calculated according to formula (4). i ': In the formula, W is the distance between the analysis camera and the tank wall away from the analysis camera, and w is the distance difference between the tank side wall near the analysis camera and the tank side wall away from the analysis camera, that is, the distance between the front side wall and the rear side wall of the tank. Since the experimental fish are located in the upper half or the lower half of the experimental area, but through calculation, it is found that the calculation formula after the calculation is the same for both cases, so no classification discussion is made here.

[0107] The upward or downward distance between frame i and frame i+1 is: Δh i =h i+1 '-h i '(5), where h i 'with h i+1 ' represents the actual swimming height of the fish from the bottom partition in frame i and frame (i+1), respectively. If Δh i> 0 is up, otherwise is down.

[0108] The average movement speed V and movement rate v from the jth second to the kth second are:

[0109]

[0110] In the formula, V represents the average movement speed from the jth second to the kth second, and if the average movement speed is positive, it indicates that the fish has an upward swimming trend from the jth second to the kth second, otherwise, it indicates a downward swimming trend.

[0111] v represents the movement rate from the jth second to the kth second, and represents the activity of the fish in the vertical direction, and the greater v is, the more active the fish is in the vertical direction, and vice versa.

[0112] Δt j-k represents the time interval, and n represents the frame rate (number of frames per second) of the camera used.

[0113] (II) is realized by a 3D behavior analysis system.

[0114] As shown in Figure 4 , the fish behavior analysis tank simultaneously meets the requirements of the fish behavior analysis tank when the camera light path is horizontal and the requirements of the fish behavior analysis tank when the camera light path is vertical, and the fish behavior analysis tank does not need to be provided with a top partition with varying thickness, as shown in Figure 5 , the left partition (water inlet partition) and the right partition (water outlet partition) are gradually thinned in the vertical camera light path direction.

[0115] A first analysis camera and a second analysis camera are respectively arranged in the vertical direction and the horizontal direction, and both analysis cameras simultaneously output corresponding auxiliary parameters (such as the second analysis camera outputs the height H i of the fish body from the tank bottom in the i-th frame image; the first analysis camera outputs the vertical line segment distance w i of the fish body from the side wall of the second analysis camera), so each direction can be analyzed independently. And according to formulas (8) to (13), the coordinates of each frame of 3D space are corrected, and the trajectory is drawn and the 3D space distribution is analyzed according to the corrected 3D coordinates of each frame of image.

[0116] The actual movement distance of the fish in the horizontal direction from the i-th frame to the i+1-th frame is x i : In the formula, a represents the thickness difference between the top end and the bottom end of the left partition or the right partition, h1 represents the distance between the first analysis camera and the tank bottom wall, h i represents the height of the partition in the camera light path direction of the first analysis camera, and l represents the distance between the bottom end of the left partition and the bottom end of the right partition, Si S represents the motion distance of the fish in the video image from the i-th frame to the i+1-th frame, i.e. the horizontal line connecting the first camera and the i-th frame and the i+1-th frame respectively and extending to the bottom of the cylinder, and the line segment formed by the two intersection points of the horizontal line with the bottom of the cylinder is S i value, x i H represents the length of the projection line of the starting point and the ending point of the actual motion trajectory on the bottom partition, i.e. the actual horizontal motion distance of the fish, H i and H i+1 respectively represent the height of the fish in the i-th frame and the i+1-th frame output by the second analysis camera.

[0117] If the frame rate of the first analysis camera is 30 frames / s, then the motion distance of the fish in the horizontal direction per second is: wherein X j is the total motion distance of the fish in the horizontal direction per second.

[0118] The actual swimming height h i ’ of the fish in the i-th frame from the bottom partition wherein W represents the distance between the second analysis camera and the cylinder wall away from the second analysis camera, w represents the distance difference between the cylinder side wall close to the second analysis camera and the cylinder side wall away from the second analysis camera, i.e. the distance between the front side wall of the cylinder and the rear side wall of the cylinder, w i represents the vertical line segment distance of the fish in the i-th frame recorded by the first analysis camera from the cylinder wall close to the second analysis camera.

[0119] The floating or sinking distance between the i-th frame and the i+1-th frame is Δh i = h i+1 ’- h i ’ (11), wherein h i ’ and h ii1 ’ respectively represent the actual swimming height of the fish in the i-th frame and the i+1-th frame from the bottom partition.

[0120] The average motion speed V and the motion rate v from the j-th second to the k-th second are:

[0121]

[0122] wherein V represents the average motion speed from the j-th second to the k-th second, if the average motion speed is positive, it indicates that the fish presents an upward swimming trend from the j-th second to the k-th second, otherwise, it indicates a downward swimming trend.

[0123] v represents the motion rate from the j-th second to the k-th second, which indicates the activity of the fish in the vertical direction, the greater v is, the more active the fish is in the vertical direction, otherwise, it is not active.

[0124] Δt j-k denotes the time interval, n represents the frame rate of the camera used.

[0125] The present application fixes the movement range of the experimental fish by setting the harmless partition assembly, and makes the direction of the partition parallel to the light path direction, avoiding the monitoring blind area generated by the conventional camera in the fish behavior analysis, so that the wall sticking movement, escape behavior and diving and bottom sticking behavior can be recorded completely and accurately. The hardware defects of the fish behavior analysis based on the camera tracking technology are compensated, and false positive or invalid results are avoided. The following is further illustrated by specific examples, test examples and comparative examples.

[0126] Example 1

[0127] This embodiment develops an analysis cylinder suitable for the Zebra Tower of the 2D zebrafish behavior observation tower of Viewpoint Company. The fish behavior analysis cylinder designed in this embodiment is made of high transparent glass, the internal size is 40*22*25cm, and the volume is 22L. It meets the size requirements proposed in embodiment 1, and the cylinder wall thickness is 0.5cm.

[0128] The partition assembly is made of acrylic material, and the design and hole distribution are as shown in Figure 2 The side view direction of the left partition (water inlet partition) and the right partition (water outlet partition) is rectangular, and the bottom of the left partition (water inlet partition) and the right partition (water outlet partition) is provided with a fecal groove, which is used to discharge fish excrement by using the hydraulic action and gravity to reduce the image recognition difficulty of zebrafish.

[0129] Since the upper part of the partition is thick and the lower part is thin, the upper part has large water flow resistance and the lower part has small resistance, and the pollutants are easy to flow into from the lower part rather than the upper part. In order to make the water inlet uniform, only the upper half of the water inlet and outlet partition is perforated. The size of the toxic area is 21*22*25cm, and the distance from the camera to the bottom of the cylinder is 92.9cm. According to formula (1), the top thickness of the water inlet and outlet partition should be 2.83cm more than the bottom thickness, and the top thickness of the front and rear partitions should be 2.96cm more than the bottom thickness. The movable area of zebrafish is fixed as a regular quadrangular pyramid by the partition assembly, the top surface and the bottom surface of the pyramid are rectangular, the top surface size is 15.34*16.03cm, the bottom surface size is 21*22cm, the pyramid height is 25cm, and the centers of the top surface and the bottom surface are located on the same vertical line.

[0130] The internal circulation filter and the pollutant dispersion device are respectively placed in the water outlet area and the water inlet area, the pollutant dispersion device includes an aeration assembly, a sewage inlet pipe and a circulation filter outlet pipe, and the water body homogenization and excrement removal are accelerated by the hydraulic flow during the operation of the two, and a heating rod is additionally arranged in the water inlet area for stabilizing the water temperature of the cultured zebrafish.

[0131] The inner loop filter is retrofitted from a miniature submersible pump, with a lift of 1.4 m and a flow rate of 1000 L / h. The water outlet is vertically upward. A microporous plate made of acrylic material is arranged on the contact surface between the pump and the water body to filter large impurities. Biochemical cotton for aquaculture is arranged outside the plate to adsorb and filter fish excrement. The plate and the biochemical cotton are fixed to the side wall of the water outlet area by a buckle made of acrylic material. A food-grade silicone tube is connected to the water outlet of the submersible pump, and the other end is connected to the water inlet area and opened 2 cm from the tube opening. A beveled PVC tube with an inner diameter of 0.2 cm is inserted into the opening, with the bevel facing the proximal tube opening. The PVC tube is connected to a food-grade silicone hose to the air, and an air volume regulating valve is installed to regulate the aeration volume. The circulating water pipe is fixed by a buckle, with the tube opening vertically downward.

[0132] Test Example 1

[0133] A fish behavior analysis tank is built according to the method in Example 1, which is suitable for the method of quantitatively evaluating behavior parameters using the Viewpoint 2D zebrafish behavior observation tower Zebra Tower:

[0134] (1) When the vertical downward shooting (i.e., the top view direction) is used as the analysis direction, the horizontal lateral shooting auxiliary camera is selected as the positioning image. Here, the camera provided with the Zebra Tower is used as the analysis camera, and the behavior parameters are quantitatively output by the Viewpoint software. A common color camera with a resolution of 30 frames / s and 1288*964 is used as the auxiliary camera to record the height data of the zebrafish. Figure 1 (2) The fixed horizontal direction auxiliary camera is located on the front of the tank, and the shooting direction is as shown in the front view position. The center of the auxiliary camera is on the central horizontal line of the zebrafish activity area, so that the video image of the zebrafish activity area in the horizontal direction is located at the center of the entire shooting area. Since the horizontal direction camera is an auxiliary camera, the distance is not strictly required, and only the position of the zebrafish needs to be identified. In this test example, in order to minimize the distortion in the horizontal direction, the auxiliary camera is set at a distance of 1.24 m from the tank wall.

[0135] Figure 1 (3) Record the fish behavior video in the vertical and horizontal shooting directions and perform trajectory recognition and behavior parameter output. Analyze the video to output the trajectory of the zebrafish and the movement distance according to the set scale. The scale is set based on the distance from the bottom of the water inlet and outlet partition, i.e., the side length of the experimental area in the video image is set to l, and the output interval is every frame. The movement distance of the fish body in the video image between the i-th frame and the i+1-th frame is set to S i . The height of the fish body from the tank bottom in the i-th frame image is output as H i .

[0136] (3) Record the fish behavior video in the vertical and horizontal shooting directions and perform trajectory recognition and behavior parameter output. Analyze the video to output the trajectory of the zebrafish and the movement distance according to the set scale. The scale is set based on the distance from the bottom of the water inlet and outlet partition, i.e., the side length of the experimental area in the video image is set to l, and the output interval is every frame. The movement distance of the fish body in the video image between the i-th frame and the i+1-th frame is set to S i . The height of the fish body from the tank bottom in the i-th frame image is output as H i .

[0137] The shooting partition will have slight arc deformation, but the deformation can be ignored if the shooting distance is sufficient, so the shooting area is a nearly conical area, and the activity of zebrafish in all directions can be calculated. The actual movement distance in the horizontal direction of the i-th frame and the i+1-th frame is x i .

[0138] Since the minimum recording time of Viewpoint output is 1s, the movement distance per second is taken as the minimum time interval here, i.e. the horizontal movement distance of the i-th second and the i+1-th second is taken as x i . The height of the frame number corresponding to the start and end time of each second output by the auxiliary camera is calculated by formula (2). Taking 1 minute of output as an example, the movement distance of the Viewpoint output is corrected by formula (2). The output results are shown in Table 1.

[0139] Table 1. Results before and after correction of movement parameters in test example 1

[0140]

[0141] Overall, the corrected movement distance is less than the directly output movement distance, which is caused by the fact that the distance from the analysis camera to the bottom of the cylinder is greater than the projection distance. The higher the height, the greater the reduction of the corrected movement distance compared to the output movement distance. The linear correlation between the reduction (reduction rate) and the height is strong (R 2 = 0.9999), which also shows that as the height of the zebrafish increases, the error of the output movement parameters compared to the actual values increases, which may produce some false positive or false negative results, such as the zebrafish moving at a normal speed at the water surface, but according to the error caused by the set scale and projection, the zebrafish is moving at a higher speed. At the same time, in a regular cubic fish tank, the zebrafish is prone to leave the monitoring interval and enter the area close to the cylinder wall and the partition at the position close to the water surface. The loss of video recognition will cause the data of the movement distance, movement times and movement time of the zebrafish to change, and there is a possibility of being identified as abnormal behavior.

[0142] Example 2

[0143] An analysis camera (resolution 3840*2160, frame rate 25 frames / s) was used for horizontal analysis, and the distance between the analysis camera and the near-end cylinder wall was 70 cm. The analysis camera was fixed on a metal frame, and a rubber non-slip pad was set at the bottom of the metal frame and placed on the same experimental table as the fish behavior analysis tank. The cylinder body was placed in the center of the frame to ensure that the fish behavior analysis tank photographed by the analysis camera was located in the center of the video image. The positioning camera used a camera with a resolution of 1280*720.

[0144] The experimental cylinder with internal dimensions of 40*25*25cm was used as the cylinder body, and the baffle assembly was designed according to Figure 3 .

[0145] The vertical dimension of the exposure area away from the analysis camera was set to 25*25cm, that is, the parallel line distance of the narrower end of the baffle was 25cm, and the height was the same as the cylinder. The relative distance from the cylinder wall in the light path direction of the camera was also 25cm. According to formula (1), the thickness difference between the top (wider side) and the bottom (narrower side) of the left baffle, the right baffle, the top baffle and the bottom baffle was 4.46cm.

[0146] The internal circulation filter was modified from a miniature submersible pump with a lift of 1.4m and a flow rate of 1000L / h. The outlet was vertically upward. A microporous plate made of acrylic was arranged on the contact surface between the pump and the water body to filter large impurities. Biochemical cotton for aquaculture was arranged outside the plate to adsorb and filter fish excrement. The plate and the biochemical cotton were fixed to the side wall of the outlet area with acrylic buckles. Food-grade silica gel tube was connected to the outlet of the submersible pump, and the other end was connected to the inlet area and opened 2cm from the tube opening. A beveled PVC pipe with an inner diameter of 0.2cm was inserted into the opening, with the bevel facing the proximal tube opening. The PVC pipe was connected to the food-grade silica gel hose to the air, and an air volume regulating valve was installed to regulate the aeration volume. The circulation pipe was fixed with buckles to make the tube opening vertically downward.

[0147] Test Example 2

[0148] Method for quantitatively evaluating behavior parameters of fish behavior analysis tank built in Example 2: (1) When the horizontal shooting is selected as the analysis direction, the auxiliary camera shooting vertically downward is selected as the positioning image.

[0149] (2) The position of the horizontal analysis camera is fixed on the horizontal line in the center of the fish activity area. W is the distance between the cylinder wall close to the analysis camera and the cylinder wall away from the analysis camera, and w is the distance difference between the cylinder side wall close to the analysis camera and the cylinder side wall away from the analysis camera. The cylinder, camera height and baffle size all need to meet the relationship described in formula (1).

[0150] (3) The vertical auxiliary camera is fixed directly above the cylinder, and the shooting center of the auxiliary camera is on the central vertical line of the fish activity area, so that the video image of the fish activity area in the vertical direction is located at the center of the whole shooting area. Since the vertical camera is an auxiliary camera, the distance is not strictly required, and only the position of the fish needs to be recognized. In order to reduce the error caused by deformation in this shooting direction, the shooting distance is increased as much as possible. In this test example, the distance between the auxiliary camera and the top of the cylinder is 1.4m.

[0151] (4) Record fish behavior video in vertical and horizontal shooting direction and perform trajectory recognition and behavior parameter output.

[0152] The trajectory of the fish and the movement distance according to the scale set are analyzed, and the scale is set according to the relative distance from the bottom (narrow side) of the partition plate away from the cylinder wall end of the analysis camera. The vertical downward auxiliary camera records the distance w of the i-th frame fish body from the vertical line segment close to the cylinder wall of the analysis camera i , and the horizontal analysis camera records the height h of the i-th frame fish body from the bottom partition plate in the image i .

[0153] The actual swimming height of the i-th frame fish body from the bottom partition plate (i.e. the inclined plane of the cylinder bottom) is calculated by formula (4). The height change of each frame is analyzed, and there are 25 frames per second, so only the analysis results of 5 seconds are shown, as shown in Table 2. The model fish in this test example is blue gourami.

[0154] Table 2 Test Example 2 Results Before and After Correcting Motion Parameters

[0155]

[0156]

[0157]

[0158] The analysis camera used is 25 frames per second, and a total of 5 seconds are analyzed. The values of h i , w i of each frame are counted, and h i’ is calculated according to formula (4), Δh i is calculated according to formula (5), and V is calculated according to formula (6). It can be seen that as w i increases, the difference between h i ' and h i becomes smaller. This method can realize the output of the real height of the test fish from the inclined plane of the cylinder bottom, especially for the bottom behavior recognition close to the side of the analysis camera. It can effectively improve the accuracy of the behavior analysis results and reduce the possibility of false negative results.

[0159] Example 3

[0160] A monitoring camera (resolution 2560x1440, frame rate 30 frames / s) is used for horizontal monitoring, and a container with the same size as in Example 1 is used. Since the horizontal shooting does not directly shoot the cylinder bottom, the fish excrement has less effect on it. The design of the inclined plane is the same as that of the light path correction partition plate. The monitoring direction is perpendicular to the 40*25cm plane, and the distance from the near-end cylinder wall is 50cm. The monitoring camera is fixed to the observation table by nuts.

[0161] The left and right side partitions are designed as shown in Figure 3 The correction of the lateral shooting does not require front and back side partitions, and the setting of the cylinder bottom wall inclined surface is equivalent to the bottom partition of the partition assembly, and only a top partition needs to be additionally added. However, due to the inconvenience of disassembly and assembly, the fixation is not firm enough, and the fish behavior on the water surface cannot be observed, etc., the top partition can be selected not to be installed. The length of all the partitions is consistent, which is 22 cm, the front side (near the camera side) of all the partitions is 3.71 cm thick, and the back side is 0.5 cm thick, and the bottom inclined surface is higher by 3.21 cm on the front side than on the back side. The left and right side partitions are provided with water inlet and outlet holes in half of the part near the camera to avoid hydraulic resistance.

[0162] The setting of the top partition will cause the water volume in the cylinder to decrease, and the water flow driving efficiency will increase due to the constraint of the partition. The parameters of the internal circulation filter are adjusted, the lift of the used miniature submersible pump is 1.2 m, the flow is 800 L / h, the water outlet is vertically upward, and a microporous hole plate made of acrylic material is arranged on the contact surface between the pump and the water body to filter large impurities. The outside of the hole plate is provided with special biochemical cotton for aquaculture for adsorption and filtration of fish excrement, the side of the hole plate and the biochemical cotton is towards the side wall of the water outlet area, and is fixed by a buckle made of acrylic material. A food-grade silicone tube is connected to the water outlet of the submersible pump, the other end is connected to the water inlet area, and is opened at a distance of 2 cm from the pipe opening of the water inlet area, and is inserted into a beveled PVC pipe with an inner diameter of 0.2 cm. The bevel faces the proximal end of the pipe opening, and the PVC pipe connects the food-grade silicone hose to the air, and installs an air volume regulating valve to adjust the aeration. The circulation water pipe is fixed by a buckle, and the pipe opening is vertically downward.

[0163] Example 4

[0164] A fish behavior analysis cylinder suitable for the Viewpoint 2D zebra fish behavior observation tower is developed. The designed analysis cylinder is made of high transparent glass material, the internal size is 50*30*15 cm, and the volume is 22.5 L. The size meets the size requirements of the technical scheme, and the cylinder wall thickness is 0.5 cm.

[0165] The partition assembly is made of acrylic material, and the design and hole arrangement are as shown in Figure 2As shown, the cross section of the left and right side partitions (water inlet and outlet partitions) along the water flow direction is a right trapezoid, the bottom hypotenuse is arranged to form a right triangle with the cylinder body, and the bottom of the water outlet partition is provided with a feces groove, the purpose of which is to use the hydraulic action and gravity to discharge fish excrement. Since the upper part of the partition is thick and the lower part is thin, contaminants are more likely to flow in from the lower part than the upper part, and in order to make the water inlet uniform, only the upper half of the water inlet and outlet partition is perforated. The size of the contaminated area is 30*30*15 cm, the camera is 40 cm away from the 30*30 cm bottom surface, the height of the partition is designed to be 15 cm, the height of the lower inclined surface is 3 cm, the thickness difference between the upper and lower parts of the left and right side partitions is calculated to be 5.06 cm using formula 1, the thickness of the lower part of the left and right side partitions is designed to be 0.5 cm, and therefore the thickness of the upper part is 5.56 cm. The thickness difference of the front side partition is 4.5 cm, the upper thickness is 5 cm, the thickness difference of the rear side partition is 5.63 cm, and the upper thickness is 6.13 cm, thus obtaining the design of the partition assembly and the cylinder body.

[0166] The inner circulating filter and the contaminant dispersion device are placed in the water outlet area and the water inlet area respectively, the contaminant dispersion device includes an aeration assembly, a sewage inlet pipe and a circulating filter outlet pipe, and the hydraulic flow during the operation of the two accelerates the homogeneity of the water body and the removal of excrement, and a heating rod is additionally arranged in the water inlet area for stabilizing the water temperature for breeding zebrafish.

[0167] The inner circulating filter is modified from a miniature submersible pump, the lift is 1.4 m, the flow is 1000 L / h, the water outlet is vertically upward, a microporous hole plate made of acrylic material is arranged on the contact surface between the pump and the water body to filter large impurities, and a special biochemical cotton for breeding is arranged outside the hole plate for adsorption and filtration of fish excrement, the hole plate and the biochemical cotton are fixed by a buckle made of acrylic material. A food-grade silica gel pipe is connected to the water outlet of the submersible pump, the other end is connected to the water inlet area, and an opening is opened at the pipe opening of the water inlet area 2 cm away from the opening, a bevelled PVC pipe is inserted, the inner diameter of the pipe is 0.2 cm, the bevel is towards the proximal opening, the PVC pipe connects the food-grade silica gel hose to the air, and an air volume regulating valve is installed to regulate the aeration amount. The circulating water pipe is fixed by a buckle, and the pipe opening is vertically downward.

[0168] Comparative Example 1

[0169] The method for quantitatively evaluating behavioral parameters of Test Example 1 was used to test the conventional partition and the improved partition described in Example 1 respectively.

[0170] The conventional partition is cuboid, with the size of 22*25*0.5cm, one side with the size of 22*0.5cm is arranged close to the bottom surface of the cylinder, and is fixed by using acrylic buckles and suction cups without bisphenol A addition. The side with the size of 22*25cm is perpendicular to the bottom surface of the cylinder, and other auxiliary devices are arranged and implemented in the fish behavior analysis cylinder of Example 1. The behavior analysis is performed by using the 2D behavior analysis system of Viewpoint Company. Since the system distinguishes between living beings and non-living beings by light transmittance, the cylinder wall connection, opaque part and water surface edge are all recognized as living beings, and the selected image area is only the bottom area of the zebrafish movable area. The experimental area surrounded by the two partition assemblies is selected in the above manner, and the motion time parameter is tested, and the recording frequency is 1 time per minute, so that the time less than 60s is not recorded.

[0171] As shown in Figure 6 , the unrecorded time of the conventional partition is 3.41%±5.34%(2.04±3.20s / min), as shown in Figure 7 , the unrecorded time of the improved partition is 1.57%±2.64%(0.94±1.59s / min), and the unrecorded time of the improved partition is shortened by 53.92% compared with the unrecorded time of the conventional partition. In this comparative example, the pre-added water in the cylinder is pure water, so the trajectory loss (wall sticking, wall collision swimming) of zebrafish is less, and most of them are cruising, which leads to less staying in the same position, so the improvement level is relatively low. In previous experiments, some pollutants (such as phenol) will enhance the escape behavior of zebrafish, and zebrafish will perform high-frequency and long-time wall collision swimming to the cylinder wall and the partition to try to escape from the high-concentration exposure area. In this case, the partition described in Example 1 can greatly reduce the loss of motion trajectory.

[0172] Of course, the above description is not a limitation of the present application, and the present application is not limited to the above examples. Changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present application should also be within the scope of the present application.

Claims

1. A fish behavior analysis tank based on improved camera technology, characterized in that, It includes a cylinder and a baffle assembly. The baffle assembly is located inside the cylinder. The inner wall of the baffle assembly is designed parallel to the camera optical path. The baffle assembly is composed of multiple baffles with gradually varying thicknesses. The internal space of the baffle assembly is the experimental area. The thickness of the baffles gradually decreases along the camera optical path. The baffle assembly includes a left baffle and a right baffle. The inner surfaces of the left baffle and the right baffle are trapezoidal in cross-section along the water flow direction. The bottom and side surfaces of the left baffle and the right baffle are adapted to fit the bottom wall of the cylinder. The bottom of the left baffle and the right baffle are provided with a manure trough. The upper middle part of the left baffle and the right baffle are provided with multiple evenly distributed inlet and outlet water passages. When the camera optical path is vertical, the partition assembly also includes a front partition and a rear partition, which are tightly fitted to the front and rear side walls of the cylinder, respectively. When the camera optical path is horizontal, the partition assembly also includes a bottom partition and a top partition.

2. The fish behavior analysis tank based on improved camera technology according to claim 1, characterized in that, The formula for calculating the thickness of the partition is: In the formula, a represents the thickness difference between the end of the partition closest to the camera and the end furthest from the camera, h1 represents the distance between the camera and the cylinder wall furthest from the camera, h2 represents the height of the partition in the direction of the camera light path, and l represents the distance between the ends of the two partitions furthest from the camera.

3. The fish behavior analysis tank based on improved camera technology according to claim 2, characterized in that, When the camera light path is vertical, the partition assembly is provided with a top cover to prevent fish from escaping the experimental area.

4. The fish behavior analysis tank based on improved camera technology according to claim 3, characterized in that, When the camera light path is horizontal, the bottom partition is the bottom wall of the cylinder.

5. The fish behavior analysis tank based on improved camera technology according to claim 1, characterized in that, It includes an inlet zone and an outlet zone. The outlet zone is equipped with an internal circulation filter, and the inlet zone is equipped with an aeration device and a wastewater dispersion system.

6. The fish behavior analysis tank based on improved camera technology according to claim 5, characterized in that, The wastewater dispersion system includes a pollutant inlet pipe and a vertically downward-arranged circulating water pipe and aeration components. The pollutant inlet pipe and the circulating water pipe are closely fitted together, and the outlet of the pollutant inlet pipe is higher than the outlet of the circulating water pipe.

7. The fish behavior analysis tank based on improved camera technology according to claim 6, characterized in that, The inner surface of the partition assembly is generally frustum-shaped.

8. The fish behavior analysis tank based on improved camera technology according to claim 7, characterized in that, The cylinder body is made of transparent glass or transparent plastic material, and the partition and top cover are both made of transparent plastic material.

9. The fish behavior analysis tank based on improved camera technology according to claim 8, characterized in that, The plastic is acrylic plastic without any added endocrine disruptors.

10. The fish behavior analysis tank based on improved camera technology according to claim 9, characterized in that, The volume of the cylinder is not less than 10L, and the length, width and height of the cylinder do not exceed 50cm.

11. A method for evaluating behavioral parameters of a fish behavior analysis tank based on improved camera technology according to any one of claims 1-10, characterized in that, This is achieved by adding an auxiliary camera perpendicular to the analysis camera to the 2D behavior analysis system; I. Analyze the camera shooting vertically and assist the camera shooting horizontally; Quantitatively assess behavioral parameters using a fish behavior analysis tank with the camera beam in a vertical direction, including the following steps: A1. Fix the vertical analysis camera directly above the cylinder and ensure that the camera's shooting center is located on the vertical line of the center of the experimental area. A2. Fix the horizontal auxiliary camera directly in front of the cylinder and ensure that the shooting center of the auxiliary camera is located on the horizontal center line of the experimental area. A3. Record fish behavior videos in both vertical and horizontal shooting directions and perform trajectory recognition and output behavior parameters; The actual horizontal movement distance of the fish in frame i and frame i+1 is x i : In the formula, 'a' represents the thickness difference between the top and bottom of the left or right partition, 'h1' represents the distance between the analysis camera and the bottom wall of the cylinder, 'h2' represents the height of the partition in the direction of the camera's optical path, 'l' represents the distance between the bottom of the left and right partitions, and 'S' represents the distance between the bottom of the left and right partitions. i S represents the distance the fish moves in the video image from frame i to frame i+1. This is calculated by connecting the camera to the positions in frames i and i+1, extending the lines to the bottom wall of the tank, and forming the line segment S at the two intersection points with the bottom wall of the tank. i Value, x i H represents the length of the line connecting the starting and ending points of the actual movement trajectory projected onto the bottom wall of the tank, i.e., the actual horizontal movement distance of the fish. i and H i+1 These are the heights of the fish in the i-th and (i+1)-th frames output by the auxiliary camera, respectively. If the camera frame rate is 30 frames per second, then the horizontal distance the fish moves per second is: In the formula, X j Let be the total horizontal distance traveled by the fish in the j-th second. II. Analyze the camera shooting horizontally and the auxiliary camera shooting vertically; Using a fish behavior analysis tank with the camera optical path horizontal, behavioral parameters are quantitatively evaluated. The analysis targets the behavioral parameters when the fish rises or sinks. These behavioral parameters include, but are not limited to, rising or sinking speed, vertical movement distance, dwell time in the upper region, dwell time in the lower region, and swimming height. The analysis includes the following steps: B1. Fix the horizontal analysis camera directly in front of the cylinder and ensure that the camera's shooting center is located on the horizontal center line of the experimental area. B2. Fix the vertical auxiliary camera directly above the cylinder and ensure that the shooting center of the auxiliary camera is located on the vertical line of the center of the experimental area. B3. Record fish behavior videos in both vertical and horizontal shooting directions and perform trajectory recognition and output behavior parameters; A vertically downward auxiliary camera records the distance w of the fish's body from the vertical segment of the tank wall closest to the analysis camera in the i-th frame. i The horizontal analysis camera records the height h of the fish's body from the bottom partition in the i-th frame of the image. i The actual swimming height h of the fish in the i-th frame from the bottom partition is calculated according to formula (4). i ': In the formula, W is the distance between the analysis camera and the cylinder wall away from the analysis camera, and w is the distance difference between the cylinder side wall close to the analysis camera and the cylinder side wall away from the analysis camera, that is, the distance between the front side wall and the rear side wall of the cylinder. The upward or downward distance between frame i and frame i+1 is: Δh i =h i+1 '-h i '(5), where h i 'with h i+1 'Represents the actual swimming height of the fish from the bottom partition in frame i and frame (i+1), respectively; The average velocity V and speed v from second j to second k are: In the formula, V represents the average speed of movement from the j-th second to the k-th second. If the average speed of movement is positive, it means that the fish tends to float upwards from the j-th second to the k-th second; otherwise, it tends to sink downwards. v represents the movement speed from the j-th second to the k-th second, indicating the fish's activity level in the vertical direction. The larger the v, the more active the fish is in the vertical direction, and vice versa. Δt j-k This represents the time interval, where n represents the frame rate of the camera used.

12. A method for evaluating behavioral parameters of a fish behavior analysis tank based on improved camera technology according to any one of claims 1-10, characterized in that, This is achieved through a 3D behavior analysis system. The fish behavior analysis tank simultaneously meets the requirements of a fish behavior analysis tank when the camera light path is horizontal and when the camera light path is vertical. The fish behavior analysis tank does not need to be equipped with a top partition with varying thickness, and both the left and right partitions gradually become thinner in the direction of the vertical camera light path. The first analysis camera and the second analysis camera are set in the vertical and horizontal directions respectively. The coordinates of the 3D space of each frame are corrected according to formulas (8) to (13). The trajectory is depicted and the 3D space distribution is analyzed based on the corrected 3D coordinates of each frame image. The actual horizontal movement distance of the fish in frame i and frame i+1 is x i : In the formula, 'a' represents the thickness difference between the top and bottom of the left or right partition, 'h1' represents the distance between the first analysis camera and the bottom wall of the cylinder, 'h2' represents the height of the partition in the direction of the camera's optical path, 'l' represents the distance between the bottom of the left and right partitions, and 'S' represents the distance between the bottom of the left and right partitions. i S represents the distance the fish moves in the video image from frame i to frame i+1. It is the line segment formed by connecting the first camera to the positions of frames i and i+1, extending to the horizontal plane at the bottom of the tank, and the two intersection points with the horizontal plane at the bottom of the tank. i Value, x i H represents the length of the line connecting the starting and ending points of the actual movement trajectory projected onto the bottom partition, i.e., the actual horizontal movement distance of the fish. i and H i+1 These are the heights of the fish in the i-th and (i+1)-th frames output by the second analysis camera, respectively. If the frame rate of the first analysis camera is 30 frames / s, then the horizontal distance the fish moves per second is: In the formula, X j Let be the total horizontal distance traveled by the fish in the j-th second; The actual swimming height h of the fish from the bottom partition in frame i i ': In the formula, W represents the distance between the second analysis camera and the cylinder wall away from the second analysis camera, and w represents the distance difference between the cylinder sidewall close to the second analysis camera and the cylinder sidewall away from the second analysis camera, that is, the distance between the front sidewall and the rear sidewall of the cylinder. i This represents the vertical distance between the fish body in the i-th frame recorded by the first analysis camera and the tank wall on the side closest to the second analysis camera; The upward or downward distance between frame i and frame i+1 is: Δh i =h i+1 '-h i '(11), where h i 'with h i+1 'Represents the actual swimming height of the fish from the bottom partition in frame i and frame (i+1), respectively; The average velocity V and speed v from second j to second k are: In the formula, V represents the average speed of movement from the j-th second to the k-th second. If the average speed of movement is positive, it means that the fish tends to float upwards from the j-th second to the k-th second; otherwise, it tends to sink downwards. v represents the movement speed from the j-th second to the k-th second, indicating the fish's activity level in the vertical direction. The larger the v, the more active the fish is in the vertical direction, and vice versa. Δt j-k This represents the time interval, where n represents the frame rate of the camera used.

Images