WO2020122875A1

WO2020122875A1 - A smart mosquito trap for capturing different types of mosquitoes into different places and the method thereof

Info

Publication number: WO2020122875A1
Application number: PCT/US2018/064899
Authority: WO
Inventors: Lun-De Liao; Arvin LIN; De-Fu Jhang; Yuh-Shiuan CHIEN
Original assignee: National Health Research Institutes; Arvin Lin
Priority date: 2018-12-11
Filing date: 2018-12-11
Publication date: 2020-06-18

Abstract

The present invention discloses a mosquito trap for capturing two or more types of mosquitoes into different chambers. The mosquito trap comprises a mosquito recognition system and a capturing mechanism. The mosquito recognition system captures image information of a mosquito and recognizes the type of the mosquito, and the capturing mechanism captures mosquitoes to different chambers with respect to the decision of the mosquito recognition system. With the present invention, live mosquitoes can be captured and differentiated to different types; additionally, the environmental values can also be recorded for further research.

Description

A Smart Mosquito Trap for Capturing Different Types of Mosquitoes into Different

Places and the Method Thereof

BACKGROUND OF THE INVENTION

Technical Field of the Invention

[0001 ] The presently described technology generally relates to an automatic insect trap device that differentiates different types of insects. In particular, this invention relates to a mosquito trap that utilizes air sucking devices, sliding guide and machine recognition algorithms.

Background

[0002] Dengue fever and malaria spread with speed in recent years, causing serious public health issues. When the catching or destroying of mosquitoes and the curing of disease is delayed, the social cost increases day by day.

[0003 ] Killing mosquitoes is the main function of mosquito traps in the general market. Since live capture of mosquitoes is not needed by the general public, the main function of most traps is to kill mosquitoes. Even after years of research, new functions are slow to be developed. However, the way to attract mosquitoes is progressing. For example, lights, such as fluorescent or colored lamps, are used to attract mosquitoes. Gases can be mixed to simulate the environment that mosquitoes like. Increasing the instrument temperature is another way to attract mosquitoes. However, methods to capture mosquitoes alive and recognize types have not been developed.

[0004] After attracting mosquitoes, most products on the market use air suction devices to suck mosquitoes, which are then killed when sucked into the fan (US 5647164A). Other devices use insecticide to kill mosquitoes that are sucked into the box, but none of these products can capture mosquitoes alive. Constant rotation of the fan and the lack of valves (US 3120075A) to stop mosquitoes from escaping is both a waste of power and efficiency, not to mention lacking the ability of recognizing different mosquito types (US 20180279598A1).

[0005] While there are many ways to attract mosquitoes (US 20070074447A1, US 3997999A, US 20130067795A1), most mosquito traps uses whether airflow or electricity to capture or directly kill the mosquitoes (US 5647164A, US 20180279598A1). These two methods, however, is not good enough for a mosquito trap to help disease control, since preserving captured mosquitoes for further research isn’t possible if electricity is used to kill mosquitoes, and capturing with is very ineffective.

[0006] We have seen the combination of rotation paired with innovative designs can achieve the goal to differentiate objects to different places (US 4245742A), and we have also seen the use of rotation to capture other insects (US 1182622A), however, there is currently no methods that uses the idea of rotation to capture and differentiate different types of insects.

[0007] Additionally, as image recognition techniques have progressed in recent years with the rise of deep learning, the importance of collecting ‘big data’ have been more important in the field of computer vision than ever. Although usual pictures such as human faces, road signs, and hand written characters are now easily obtained with great resource to be found online, images that presents detailed characteristics of insects have been lacking, thus limiting the development of image recognition specializing in recognizing insects.

[0008] This problem became significant when we realize that some kinds of insects are hugely responsible for certain public health issues. For example, dengue fever outbreaks are almost always caused by two specific types of mosquitoes: Aedes Aegypti and Aedes Albopictus; malaria is spread by marsh mosquitoes, another type of mosquitoes responsible for countless human lives. With the current breakthrough in image recognition, people starts to apply these recognition systems into tracking the spread of certain insects in the hope to control or even prevent most of the insect-borne diseases, with that said, the lack of database in images of these insects starts to become a limiting factor.

[0009] Mosquitoes that are cultivated in the research lab are different from wild mosquitoes due to differences in environment. Blood, liquid, and viruses obtained from captured live mosquitoes will better reflect wild mosquitoes.

[0010] Thus, the current objective is to design a small and efficient smart mosquito trap that can capture and differentiate different types of live mosquitoes and simultaneously record environmental values for more efficient research.

SUMMARY OF INVENTION

[0011 ] Accordingly, embodiments of the present invention provide mosquito traps and methods for capturing two or more types of mosquitoes into different chambers. With the present invention, live mosquitoes can be captured and differentiated to different types. In addition, the environmental values can also be recorded for further research.

[0012] In order to achieve the objections above, the present invention discloses a mosquito trap for capturing two or more types of mosquitoes into different chambers. The mosquito trap comprises a mosquito recognition system and a capturing mechanism. The mosquito recognition system captures image information of a mosquito and recognizes the type of the mosquito, and the capturing mechanism captures mosquitoes to different chambers with respect to the decision of the mosquito recognition system. [0013 ] The capturing mechanism comprises a capturing box, at least one check valve, at least one channel, and at least one air sucking device. The at least one check valve has a front side and a back side. In addition, the at least one channel is placed behind each check valve. Furthermore, the at least one air sucking device, is connected at the other end of each channel. Wherein one of the air sucking devices is turned on or turned off due to the decision of the mosquito recognition system when the mosquito flies into a recognition area.

[0014] In one embodiment, each check valve comprises a back plate, a rotating shaft and an outer restricting frame. Wherein the back plate blocks air and other things, such as a mosquito, from passing by. In addition, the rotating shaft is a cylindrical pin that fits to a cylindrical hole of the back plate. Furthermore, the outer restricting frame has two cylindrical holes on its side to attach the rotating shaft. Wherein the back plate is only allowed to rotate to the back side.

[0015] In one embodiment, the channel is extended from the back side of the check valve, and the check valve is only allowed to open in the direction towards the channel.

[0016] In one embodiment, the channel is also used as a chamber to store the captured mosquitoes.

[0017] In one embodiment, the mosquito trap further comprises at least one storage area, which is placed between each channel and each air sucking device, and is used as a chamber to store the captured mosquitoes.

[0018] In one embodiment, the air sucking device generates an air flow in the channel to open the check valve when the air sucking device is turned on.

[0019] In one embodiment, the air flow sucks the mosquito into the channel which the air sucking device is turned on. [0020] In one embodiment, the air sucking device is then turned off after the mosquito is captured.

[0021 ] In one embodiment, the mosquito trap further comprises a first sensor used as an environmental sensor.

[0022] In one embodiment, the first sensor records C02 concentration when the mosquito is captured.

[ 0023 ] In one embodiment, the first sensor is a MG-811 sensor.

[0024] In one embodiment, the mosquito trap further comprises a second sensor used as an environmental sensor.

[0025] In one embodiment, the second sensor records humidity and temperature when the mosquito is captured.

[0026] In one embodiment, the second sensor is a DHT-22 sensor.

[0027 ] In one embodiment, the mosquito trap further comprises a micro-controller that sends data from the first sensor or the second sensor to a server.

[0028] In one embodiment, the mosquito trap further comprises a micro-controller that runs the mosquito recognition system.

[0029] In one embodiment, the mosquito trap further comprises an LCD screen to present the amount or types of the mosquitoes captured.

[0030] In one embodiment, the mosquito recognition system comprises a short range camera and an image recognition algorithm. Wherein the short range camera is located to focus on the capturing box, and the image recognition algorithm uses a video feed provided by the short range camera to decide whether the mosquito is in the capturing box or not. [0031 ] In one embodiment, the image recognition algorithm decides the type of the mosquito in the capturing box.

[0032] Furthermore, the present invention also discloses a method for capturing two or more types of mosquitoes into different chambers. The method includes steps of capturing image information of a mosquito and recognizing the type of the mosquito; turning on an air sucking device based on the image information and the recognition; generating an air flow in a channel to open a check valve when the air sucking device is turned on; sucking the mosquito into the channel which the air sucking device is turned on; and turning off the air sucking device after the mosquito is captured.

[0033 ] In one embodiment, the method further comprises detecting and recording environmental data.

[0034] In one embodiment, recording environmental data includes recording C02 concentration with a first sensor when the mosquito is captured.

[ 0035 ] In one embodiment, the first sensor is a MG-811 sensor.

[0036] In one embodiment, recording environmental data includes recording humidity and temperature with a second sensor when the mosquito is captured.

[0037 ] In one embodiment, the second sensor is a DHT-22 sensor.

[0038] In one embodiment, the method further comprises sending data from the first sensor or the second sensor to a server.

[0039] In one embodiment, the method further comprises presenting the amount or types of the mosquitoes captured with an LCD screen.

[0040] In one embodiment, the method further comprises providing a video feed with a short range camera. [0041 ] In one embodiment, the method further comprises using the video feed to decide whether the mosquito is in the capturing box or not.

[0042] In one embodiment, the method further comprises deciding the type of the mosquito in the capturing box.

BRIEF DESCRIPTION OF THE DRAWINGS

[0043 ] FIG. 1 is a perspective view of an air sucking mosquito trap.

[0044] FIG. 2Ais an illustration of an air sucking mosquito trap.

[0045] FIG. 2B is a disassembly diagram of FIG. 2A.

[0046] FIGS. 3A-3B are exemplary diagrams of a check valve of the air sucking mosquito trap.

[0047 ] FIG. 3C is an illustration of a check valve.

[0048] FIG. 3D is a disassembly diagram of FIG. 3C.

[0049] FIG. 4Ais an exemplary diagram of an air sucking mosquito trap.

[0050] FIG. 4B is a disassembly diagram of FIG. 4A.

[0051 ] FIG. 5 A is an illustration of a short range camera, a capturing box and check valves.

[0052] FIG. 5B is a disassembly diagram of FIG. 5 A.

[ 0053 ] FIG. 6A is an exemplary diagram of a rotating mosquito trap.

[0054] FIG. 6B is an illustration of a rotating mosquito trap.

[0055] FIG. 6C is a disassembly diagram of FIG. 6B.

[0056] FIG. 7 is an exemplary diagram of a sliding guide.

[ 0057 ] FIG. 8 A is an embodiment of a rotating mosquito trap.

[0058] FIG. 8B is a disassembly diagram of FIG. 8A. [0059] FIG. 9Ais an illustration of a mosquito imager.

[0060] FIG. 9B is a disassembly diagram of FIG. 9A.

DETAILED DESCRIPTION OF THE INVENTION

[0061 ] The spreading of mosquito borne diseases such as Dengue fever, Zika virus infections have grown widely in recent years. An important aspect of controlling and preventing these diseases is to target Aedes Aegypti and Aedes Albopictus, two types of mosquito responsible for carrying these viruses, and capture live samples of these mosquitoes for analysis. This function, however, is not implemented in most modern mosquito traps.

[0062] Accordingly, the present invention provides a mosquito trap for capturing two or more types of mosquitoes into different chambers. As shown in FIG. 1, when a mosquito 16 flies into a recognition area 18, the present invention captures image information of the mosquito 16 and recognizes the type of the mosquito 16. After the recognition, the present invention captures the mosquito 16 based on the image information and the recognition. With the present invention, live mosquitoes can be captured and classified for analysis. For example, Aedes and non-Aedes can be captured and classified by the present invention.

[0063 ] Referring to FIGS. 2A-2B and 4A-4B, an air sucking mosquito trap 10 that enables capturing different types of mosquitoes 16 into different chambers comprises a mosquito recognition system 12 and a capturing mechanism 14. The mosquito recognition system 12 captures image information of a mosquito 16 and recognizes the type of the mosquito 16, and the capturing mechanism 14 captures mosquitoes 16 to different chambers with respect to the decision of the mosquito recognition system 12. [0064] The capturing mechanism 14 comprises a capturing box 26, at least one check valve 42, at least one channel 44, and at least one air sucking device 46. The at least one channel 44 is placed behind each check valve 42, and the at least one air sucking device 46 is connected at the other end of the each channel 44. In one embodiment of the air sucking mosquito trap 10, the capturing box 26 is a box that has its top side opened (to allow mosquitoes 16 to fly through), its front side transparent (to allow the short range camera 22 to monitor the capturing box 26), its back side non-transparent, and the other two sides installed with the check valve 42 paired with the channel 44 and the air sucking device 46. In one embodiment, the channel 44 is also used as a chamber to store the captured mosquitoes 16. In one embodiment, the capturing mechanism 14 further comprises at least one storage area 48, which is placed between each channel 44 and each air sucking device 46, and is used as a chamber to store the captured mosquitoes 16.

[0065] When a mosquito 16 flies into a recognition area 18, the mosquito recognition system 12 captures image information of the mosquito 16 and recognizes the type of the mosquito 16. After the recognition, one of the air sucking device 46 is turned on based on the image information and the recognition. With the turned-on air sucking device 46, an air flow in the corresponding channel 44 is generated to open the corresponding check valve 42 and suck the mosquito 16 into the corresponding channel 44. Specifically, the generated air flow only opens the corresponding check valve 42, while other check valves 42 would remain closed. After the mosquito 16 is captured and sucked into the corresponding chamber, the turned-on air sucking device 46 is then being turned off. Therefore, the corresponding check valve 42 is then pulled down by gravity and closed, and the captured mosquito 16 is then trapped in the corresponding channel 44 or the corresponding storage area 48. [0066] Referring to FIGS. 3A-3D, each check valve 42 comprises a back plate 52, a rotating shaft 54, and an outer restricting frame 56. The back plate 52 blocks air and other things, such as a mosquito 16, from passing by. Specifically, the back plate is a plate like object with a cylindrical hole on the top for it to connect to the rotating shaft 54. The rotating shaft 54 is a cylindrical pin that fits to the cylindrical hole of the back plate 52. Once connected, the rotating shaft 54 allows the back plate 52 to rotate along it. The outer restricting frame 56 looks like any typical picture frame, except that the middle hollow area should have the shape of the back plate 52. It has two cylindrical holes on its side to attach the rotating shaft 54.

[0067 ] The check valve 42 has a front side and a back side. On the front side, the outer restricting frame 56 has a protrusion that is like a door threshold which has the purpose of blocking the back plate 52 from rotating when all parts are assembled; therefore, the back plate 52 is only allowed to rotate to the back side of the check valve 42 but not to the front side of it. When the back plate is rotated and object, such as the mosquito 16, are able to pass through the hollow area of the outer restricting frame 56, we say the check valve 42 is opened. When the back plate blocks the hollow area of the outer restricting frame 56, we say the check valve 42 is closed. Specifically, the channel 44 of the air sucking mosquito trap 10 is extended from the back side of the check valve 42, and the check valve 42 is only allowed to open in the direction towards the channel 44.

[0068] Referring to FIGS. 4A-4B and 5A-5B, the mosquito recognition system 12 comprises a short range camera 22 and an image recognition algorithm 24. The short range camera 22 is located to focus on the capturing box 26, and the image recognition algorithm 24 uses a video feed provided by the short range camera 22 to decide whether the mosquito 16 is in the capturing box 26 and the type of the mosquito 16 in the capturing box 26 In one embodiment of the present invention, an LCD screen 78 presents the amount or types of the mosquitoes 16 which are captured.

[0069] In one embodiment of the present invention, the air sucking mosquito trap 10 comprises a micro-controller 76, such as raspberry pi 3, to run the mosquito recognition system 12 In one embodiment, the image recognition algorithm 24 achieves this goal with the following steps:

[0070] Step 1, Movement Detection: Say at any time point n, the image send by the video feed is image I_n , movement detection is done by performing Gaussian blur on image I_n (this step is done to eliminate possible noises) to get image IG, we then calculate the differences of image ¾ and image I_baseiine (described in step 2) pixel by pixel to get image Imovement· With image Imovement, we then threshold it with a certain value and calculate its contours C. With every contour in C, if the contour area is in the targeted range (about the size of a mosquito), we then take a sub-image that includes the contour and feed the sub-image to the neural network (described in step 3) for recognition.

[0071 ] Step 2, Image Ibaseiine: For any time point n, the image Ibaseiine is calculated by taking the images from time point n-1, n-2, . .,ΐo n-k (k is a hyperparameter), performing Gaussian blur on every image, and taking the average of it. Image I_baseiine represents the non-moving background in the video from time point n-1 through n-k. When the difference between image I_baseiine and image I_n is taken in step 1, the difference represents the movements that happened in time point n.

[0072] Step 3, Neural Network for Recognizing Mosquitoes: When a sub-image from step 1 is given to recognize, the image is first resized into size 227x227x3, it is then feed into the neural network to predict whether the sub-image contains Aedes, Culex, or it is an empty picture. The result from the neural network is then used to control the capturing mechanism

14

[0073 ] Step 4: The neural network in step 3 uses SqueezeNet structure and is trained using about a hundred thousand labeled pictures to have the ability to recognize different types of mosquitoes.

[0074] In one embodiment of the present invention, the air sucking mosquito trap 10 comprises a first sensor 72 and/or a second sensor 74 used to detect and record the environmental data. For example, the first sensor 72 records C02 concentration while the second sensor 74 records humidity and temperature when the mosquito 16 is captured. In one embodiment, the first sensor 72 is a MG-811 sensor, and the second sensor 74 is a DHT-22 sensor.

[0075] In one embodiment of the present invention, the micro-controller 76 sends data from the first sensor 72 or the second sensor 74 to a server 92.

[0076] Referring to FIGS. 6A-6C, a rotating mosquito trap 110 that enables capturing different types of mosquitoes 116 into different chambers comprises a mosquito recognition system 112 and a capturing mechanism 114. The mosquito recognition system 112 captures image information of a mosquito 116 and recognizes the type of the mosquito 116, and the capturing mechanism 114 captures mosquitoes 116 to different chambers with respect to the decision of the mosquito recognition system 112.

[0077] The capturing mechanism 114 comprises a capturing plate 142, at least one sliding guide 144, a bottom plate 146, a motor 152, a capturing box 154, a camera holder 156 and an LED holder 158. [0078] The capturing plate 142 is a cylinder. The cylinder has a front surface and a back surface which is both circles. On the front surface, a rectangular track 162 is carved out of it, and the center of the rectangle track 162 should match the center of the front surface circle. The rectangle track 162 is carved through the entire diameter of the front surface circle but doesn’t carve through the cylinder.

[0079] In one embodiment of the rotating mosquito trap 110, the at least one sliding guide 144 looks like a train rail or other guides that could be seen in machinery designs. The at least one sliding guide 144 is attached to the rectangular track 162.

[0080] Referring to FIG. 7, the sliding guide 144 is implemented as follows:

[0081 ] Along the capturing plate’s 142 rectangular track’s 162 side, we further carve out two sub -rectangular tracks 164, these two sub -rectangular tracks 164 however doesn’t pass all the way through. A narrow straight gap is used to connect the sub -rectangular tracks 164 with the main rectangular track 162. The two sub -rectangular tracks 164 serve the same purpose as a train rail in that it guides the direction of movement of the whole sliding guide 144

[0082] The other part of the sliding guide 144 implementation is a‘sliding guide car 166’ that can move along the sub-rectangular tracks 164. The sliding guide car 166 consists of a rectangular body that fits the main rectangular track 162, and it has two wheels attached on either side to allow it to move in the sub -rectangular track 164.

[0083 ] Referring to FIGS. 6A-6C, the bottom plate 146 is attached to the sliding guide car 166. The bottom plate 146, when attached, should cover the main rectangular track 162. In addition, the motor 152 is connected to the capturing plate 142 through its shaft. The motor 152 rotates the capturing plate 142 along its center. [0084] In one embodiment, capturing box 154 is made by the following steps:

[0085] Step 1 : First design a solid rectangular box that has a width and length larger than the diameter of the front surface circle of the capturing plate 142, and the height of this rectangular box would be the same as the height of the capturing plate’s 142 cylinder.

[0086] Step 2: At the upper part of the rectangular box in step 1, carve out all the way through a cylinder with the same size (or a little larger) as the capturing plate 142 so that the capturing plate 142 could fit inside it.

[0087] Step 3: At the bottom part, carve partially two rectangular empty boxes that connects to the empty cylinder in step 2, the two boxes separates in the middle of the solid rectangular box in step 1.

[0088] Step 4: Above the solid rectangular box in step 1, where the two empty boxes in step 3 are visible, attach a transparent acrylic plate that covers the empty boxes and the empty cylinder in step 2.

[0089] Step 5: Below the solid rectangular box in step 1, where the two empty boxes in step 3 are not visible. Attach a motor holder that holds the motor 152. The motor holder is a plate that is big enough to cover the empty circle carved out in step 2, and it has a hollow hole in the middle to allow the motor’s 152 shaft to pass through and connects to the capturing plate 142.

[0090] In one embodiment, the camera holder 156 holds a short range camera 168 and allows it to aim at a capturing area 126. In addition, the LED holder 158 holds a LED to provide lighting to the capturing area 126. Furthermore, the LED holder 158 is attached above the camera holder 156. [0091 ] When the rotating mosquito trap 110 is at rest, the empty rectangular track 162 in the capturing plate 142 aligns with the opening of the capturing box 154. The bottom plate 146 is at its lowest limit, thus forming a space in the rectangular track 162 between the bottom plate 146 and the opening of the capturing box 154. This area is called the capturing area 126.

[0092] When a mosquito 116 is detected in the capturing area 126, the capturing plate 142 is then rotated by the motor 152, which immediately closes the gap and traps the mosquito 116 in the capturing area 126. When the capturing plate 142 rotates over 90 degrees, the bottom plate 146 is then pulled down by gravity. When the bottom plate 146 starts gliding along the sliding guide 144, the space of the capturing area 126 becomes smaller and forces the mosquito 116 to move outwards in the capturing plate 142.

[0093 ] Finally, the bottom plate 146 reaches its upper limit. There is no space left in the capturing area 126, and the mosquito 116 is forced to be in a storage area 148 in the capturing box 154. After capturing the mosquito 116, the capturing plate 142 rotates the other way back to its original position. The bottom plate 146 is then again pulled by gravity and gets to its lowest limit. The capturing area’s 126 space is restored and ready to catch another mosquito 116.

[0094] Since the capturing mechanism 114 requires rotating the capturing plate 142 to at most 180 degrees, we can control the capturing plate 142 to rotate in different directions when capturing different types of mosquitoes 116, thus achieving the goal to effectively capture different types of mosquitoes 116 into different areas.

[0095] In one embodiment, the mosquito recognition system 112 comprises a short range camera 168 and an image recognition algorithm 182. The short range camera 168 is located to focus on the capturing area 126, and the image recognition algorithm 182 uses a video feed provided by the short range camera 168 to decide whether the mosquito 116 is in the capturing area 126 and the type of the mosquito 116 in the capturing area 126. In one embodiment of the present invention, an LCD screen 178 presents the amount or types of the mosquitoes 116 which are captured.

[0096] In one embodiment, the rotating mosquito trap 110 comprises a micro-controller 176, such as raspberry pi 3, to run the mosquito recognition system 112. In one embodiment, the image recognition algorithm 182 achieves this goal with the following steps:

[0097] Step 1, Movement Detection: Say at any time point n, the image send by the video feed is image I_n , movement detection is done by performing Gaussian blur on image I_n (this step is done to eliminate possible noises) to get image L;, we then calculate the differences of image IG and image I_baseiine (described in step 2) pixel by pixel to get image Imovement· With image Imovement, we then threshold it with a certain value and calculate its contours C. With every contour in C, if the contour area is in the targeted range (about the size of a mosquito), we then take a sub-image that includes the contour and feed the sub-image to the neural network (described in step 3) for recognition.

[0098] Step 2, Image Ibaseiine: For any time point n, the image Ibaseiine is calculated by taking the images from time point n-1, n-2, . .,ΐo n-k (k is a hyperparameter), performing Gaussian blur on every image, and taking the average of it. Image I_baseiine represents the non-moving background in the video from time point n-1 through n-k. When the difference between image I_baseiine and image I_n is taken in step 1, the difference represents the movements that happened in time point n. [0099] Step 3, Neural Network for Recognizing Mosquitoes: When a sub-image from step 1 is given to recognize, the image is first resized into size 227x227x3, it is then feed into the neural network to predict whether the sub-image contains Aedes, Culex, or it is an empty picture. The result from the neural network is then used to control the capturing mechanism

114

[00100] Step 4: The neural network in step 3 uses SqueezeNet structure and is trained using about a hundred thousand labeled pictures to have the ability to recognize different types of mosquitoes.

[00101 ] In one embodiment, the rotating mosquito trap 110 comprises a first sensor 172 and/or a second sensor 174 used to detect and record the environmental data. For example, the first sensor 172 records C02 concentration while the second sensor 174 records humidity and temperature when the mosquito 116 is captured. In one embodiment, the first sensor 172 is a MG-811 sensor, and the second sensor 174 is a DHT-22 sensor.

[00102] In one embodiment, a box 184 contains the micro-controller 176. In one embodiment, the micro-controller 176 collects data from the first sensor 172 or the second sensor 174 and sends it to a server 192.

[00103 ] Referring to FIGS. 8A-8B, the rotating mosquito trap 110 is implemented with an outer box 186, which comprises a base container 194, a top container 196 and a top cover

198

[00104] In one embodiment, the base container 194 is like a bucket, and it has several openings to allow water or other material to be filled inside this base container 194. In addition, the top container 196 has the shape similar to the base container 194; however, it has several holes to allow any kind of gas to pass through at the bottom. The capturing mechanism 114 could be placed on the top container 196. The top cover 198 covers the top container 196, but leaves a hole in the middle to connect to the top of the capturing area 126.

[00105] After assembly, the whole rotating mosquito trap 110 will look like FIG. 8 A. When water is filled in the base container 194, water would evaporate and escape through the hole in the top cover 198, luring mosquitoes 116 to fly in. Once a mosquito 116 flies inside, it would be recognized by the mosquito recognition system 112, and being captured by the capturing mechanism 114.

[00106] Moreover, a mosquito imager 200 is designed to take images of any given mosquito in all 3 -dimension orientations possible to obtain a huge data set of mosquitoes with wide variety. The data set obtained is essential for training the mosquito recognition system 12 utilized in the air sucking mosquito trap 10 and the mosquito recognition system 112 utilized in the rotating mosquito trap 110. The mosquito imager 200 is thus a design that aids the development of the mosquito recognition system 12 in the air sucking mosquito trap 10 and the mosquito recognition system 112 in the rotating mosquito trap 110.

[00107] Referring to FIGS. 9A-9B, the mosquito imager 200 that takes pictures from all angles for data generation needed for image recognition comprises an insect pin 210, a base holder 220, a base motor 230, a base shelf 240, a camera motor 250, a camera platform 260, a camera 270, an LED 280 and position sensors 290. Wherein the insect pin 210 is used in the creation of insect specimens, and the specimens are fixed on the tip of the insect pin 210

[00108] The base holder 220 constitutes a thin cylinder. At both surfaces of the cylinder, a circular plate is attached. At one side, the circular plate has a motor connector at the center that can be attached to the base motor 230. The circular plate at the other side has a slight hole that is used to attach the insect pin 210. The slight hole is perpendicular to the surface and sits in the center of the plate.

[00109] A base motor 230, such as a stepper motor, can rotate the base holder 220, and thus rotating the specimen that sits on top with high precision with every step.

[00110] Abase shelf 240 is used to hold the base motor 230, the camera motor 250, and the camera platform 260. The base shelf 240 comprises a base rectangular plate 242, a top rectangular plate 244, a motor side plate 246 and a normal side plate 248. Wherein the base motor 230 sits on the base rectangular plate 242, and the top rectangular plate 244 is on top of the base motor 230. In addition, the top rectangular plate 244 has the same size as the base rectangular plate 242, with a circular hole in the center to allow the base motor’s 230 shaft to extrude from it. Moreover, the motor side plate 246 is a rectangular plate that being placed at and attached to the side of the base rectangular plate 242 and the top rectangular plate 244. On the top of the motor side plate 246, a hole is made for the camera motor’s 250 shaft to extrude from it. The height of the hole is set such that when the camera motor 250 is attached, it’s axis of rotation would pass through the tip of the insect pin 210, thus every object, such as the mosquito, which is rotated by the camera motor 250 would also rotate around the tip of the insect pin 210. Furthermore, the normal side plate 248 is a rectangular side plate that being placed at and attached to the other side of the base rectangular plate 242 and the top rectangular plate 244. On the top of the normal side plate 248, a mechanical bearing is attached in the position that its rotation axis matches the rotation axis of the camera motor

250 [00111 ] The camera motor 250, such as a stepper motor, is placed and fixed on the motor side plate 246 in the base shelf 240. The camera motor 250 can rotate the camera platform 260 with high precision.

[00112] The camera platform 260 comprises a camera plate holder 262 and two support sticks 264. Wherein the camera plate holder 262 is a flat plate that being used to hold the camera 270 and the LED 280. In addition, the two support sticks 264 would be placed at both sides of the camera plate holder 262. One support stick 264 would have a motor connector that connects it with the camera motor 250, the other would have a mechanical bearing connector that connects it to the mechanical bearing in the normal side plate 248 of the base shelf 240.

[00113 ] The camera 270 is used to take pictures. In one embodiment, the camera 270 is placed on the camera plate holder 262. Once placed, the tip of the insect pin 210 would appear at the center of the image the camera 270 would capture. As the camera motor 250 rotates, the camera 270 would take images of the tip of the insect pin 210 at different angles, but the distance between the camera 270 and the tip of the insect pin 210 would always remain the same, and the tip would always remain in the center of the image taken.

[00114] The LED 280 is on the camera plate holder 262. In one embodiment, several LEDs 280 are placed on the camera plate holder 262 to control the lighting conditions of the image taken.

[00115] In one embodiment, several position sensors 290 are placed to detect the position of the camera motor 250, and thus ensuring the camera 270 is placed at the desired location. In one embodiment, one position sensor 290 is placed to detect the position of the base holder 220, and thus ensuring the base holder 220 is at the right position. [00116] With the mosquito imager 200, we can generate images of a specimen in almost all angles by rotating either the camera motor 250 or the base motor 230. The mosquito imager 202 is used to generate data that is used to train a neural network for the air sucking mosquito trap 10 and the rotating mosquito trap 110; however, the mosquito imager 200 is suitable to take images of any kind of insect with similar sizes.

Claims

1. A mosquito trap for capturing two or more types of mosquitoes into different chambers, the mosquito trap comprising:

a mosquito recognition system, which captures image information of a mosquito and recognizes the type of the mosquito; and

a capturing mechanism, which captures mosquitoes to different chambers with respect to the decision of the mosquito recognition system, the capturing mechanism comprising: a capturing box;

at least one check valve, which has a front side and a back side;

at least one channel, which is placed behind each check valve; and

at least one air sucking device, which is connected at the other end of each channel; wherein one of the air sucking devices is turned on or turned off due to the decision of the mosquito recognition system when the mosquito flies into a recognition area.

2. The mosquito trap of claim 1, wherein each check valve comprises:

a back plate, which blocks air and other things, such as a mosquito, from passing by;

a rotating shaft, which is a cylindrical pin that fits to a cylindrical hole of the back plate; and

an outer restricting frame, which has two cylindrical holes on its side to attach the rotating shaft;

wherein the back plate is only allowed to rotate to the back side.

3. The mosquito trap of claim 1, wherein the channel is extended from the back side of the check valve, and the check valve is only allowed to open in the direction towards the channel.

4. The mosquito trap of claim 1, wherein the channel is also used as a chamber to store the captured mosquitoes.

5. The mosquito trap of claim 1, further comprising at least one storage area, which is placed between each channel and each air sucking device, and is used as a chamber to store the captured mosquitoes.

6. The mosquito trap of claim 1, wherein the air sucking device generates an air flow in the channel to open the check valve when the air sucking device is turned on.

7. The mosquito trap of claim 6, wherein the air flow sucks the mosquito into the channel which the air sucking device is turned on.

8. The mosquito trap of claim 7, wherein the air sucking device is then turned off after the mosquito is captured.

9. The mosquito trap of claim 1, further comprising a first sensor used as an environmental sensor.

10. The mosquito trap of claim 9, wherein the first sensor records C02 concentration when the mosquito is captured.

11. The mosquito trap of claim 10, wherein the first sensor is a MG-811 sensor.

12. The mosquito trap of claim 1, further comprising a second sensor used as an environmental sensor.

13. The mosquito trap of claim 12, wherein the second sensor records humidity and temperature when the mosquito is captured.

14. The mosquito trap of claim 13, wherein the second sensor is a DHT-22 sensor.

15. The mosquito trap of claim 9-14, further comprising a micro-controller that sends data from the first sensor or the second sensor to a server.

16. The mosquito trap of claim 1, further comprising a micro-controller that runs the mosquito recognition system.

17. The mosquito trap of claim 1, further comprising an LCD screen to present the amount or types of the mosquitoes captured.

18. The mosquito trap of claim 1-8, wherein the mosquito recognition system comprises: a short range camera, which is located to focus on the capturing box; and

an image recognition algorithm, which uses a video feed provided by the short range camera to decide whether the mosquito is in the capturing box or not.

19. The mosquito trap of claim 18, wherein the image recognition algorithm decides the type of the mosquito in the capturing box.

20. The mosquito trap of claim 18, further comprising a first sensor and a second sensor used as environmental sensors.

21. The mosquito trap of claim 20, wherein the first sensor records C02 concentration and the second sensor records humidity and temperature when the mosquito is captured.

22. The mosquito trap of claim 21, further comprising a micro-controller that runs the mosquito recognition system and sends data from the first sensor or the second sensor to a server.

23. The mosquito trap of claim 21, wherein the first sensor is a MG-811 sensor and the second sensor is a DHT-22 sensor.

24. A method for capturing two or more types of mosquitoes into different chambers, the method comprising:

capturing image information of a mosquito and recognizing the type of the mosquito; turning on an air sucking device based on the image information and the recognition; generating an air flow in a channel to open a check valve when the air sucking device is turned on;

sucking the mosquito into the channel which the air sucking device is turned on; and turning off the air sucking device after the mosquito is captured.

25. The method of claim 24, further comprising detecting and recording environmental data.

26. The method of claim 25, wherein recording environmental data includes recording C02 concentration with a first sensor when the mosquito is captured.

27. The method of claim 26, wherein the first sensor is a MG-811 sensor.

28. The method of claim 25, wherein recording environmental data includes recording humidity and temperature with a second sensor when the mosquito is captured.

29. The method of claim 28, wherein the second sensor is a DHT-22 sensor.

30. The method of claim 26-29, further comprising sending data from the first sensor or the second sensor to a server.

31. The method of claim 24, further comprising presenting the amount or types of the mosquitoes captured with an LCD screen.

32. The method of claim 24, further comprising providing a video feed with a short range camera.

33. The method of claim 32, further comprising using the video feed to decide whether the mosquito is in the capturing box or not.

34. The method of claim 33, further comprising deciding the type of the mosquito in the capturing box.

35. The method of claim 33, further comprising detecting and recording environmental data.

36. The method of claim 35, wherein recording environmental data includes recording C02 concentration with a first sensor and recording humidity and temperature with a second sensor when the mosquito is captured.

37. The method of claim 36, further comprising sending data from the first sensor or the second sensor to a server.

38. The method of claim 36, wherein the first sensor is a MG-811 sensor and the second sensor is a DHT-22 sensor.