Description of the Prior Art Fish farming is a growing industry, mainly due to an increasing demand for seafood, which can only be responded by an increase in fish farming. One type of fish farming is a fish farming in pens. One pen typically comprises circular or rectangular floating and a net.

Surveying pens is important, mainly due to loss of fish when the net is ruptured and also due to the risk that the cultivated fish mingles with wild fish. There is therefore an increasing demand for intensive surveillance to avoid this. Today a single pen typically produces around 8000kg of fish annually. The importance of keeping such pens intact is therefore high.

The fish farmer has two options for surveying the pens. First by using divers, which usually are required to work at least two together. This is however very expensive, time- consuming and also not a very efficient way of surveying when the area to be inspected is large. The second way is to use a cable-connected camera driven with a motor (remotely operated vehicle, ROV), which is guided by an operator. This means that at least one person has to be present in a boat to control the camera. The advantage of this method is that no divers are needed. The disadvantage on the other hand is the manpower that is needed and equipment, besides the coverage is also limited and very dependent on the weather conditions.

General description of the invention It is an object of the present invention to automatically survey fish farming pens nearly continuously and with more accuracy then presently possible and at the same time reduce the cost of such a process.

It is another object of the present invention to provide a method and a system for monitoring the environmental parameters in the surroundings of the underwater pens.

According to a first aspect, the present invention relates to a method for underwater surveillance by means of using an autonomous underwater vehicle (AUV), the method comprising the steps of: moving the AUV in an area to be inspected, wherein images of object areas are frequently captured by the AUV and stored, locating the AUV with respect to a predetermined coordinate system, and associating coordinates of the AUV at any instant of time to the captured images at the same instant of time.

The motion can be predefined and based on geometrical information in said coordinate system regarding the area to be inspected, wherein the geometrical information can comprise a sequence of coordinates in said coordinate system that the AUV follows. The geometrical information can be downloaded by an operator to the AUV through a communication channel, which is preferably wireless. The communication channel can for example be a telephone connection or the Internet.

The captured images can be captured at a predefined time interval or the number of images obtained by the AUV and/or their location can be predetermined, where the capturing means can be a sonar or digital camera. The location of the AUV can be performed by means of using a long baseline system consisting of a plurality of baseline stations, wherein the distance from the AUV to the plurality of external baseline stations is calculated by using time measurement and based on the calculated distance the coordinate of the AUV with respect to the coordinate system are determined. The acoustic baseline stations can periodically emit a sound wave or they can be activated through a sound wave emitted by the AUV. The baseline stations can for example be three with a fixed known position and/or fixed internal distance, located at the surface of the water, at a fixed height in the water or at the

bottom of the water. The location of the AUV can also be based on dead reckoning using a Doppler velocity log and compass, with or without inertial navigation system.

The captured images can be processed in the AUV or externally after the AUV has finished said predefined path either automatically or visually by the operator. When processed in the AUV and the result is a deviation from the normal state, wherein the normal state can be a reference parameter, the AUV sends an alarm signal by transmitting sound wave, with for example an acoustic modem, or by means of moving the AUV to a docking station where operators are warned. After the AUV has finished said predefined path it returns to the docking station and uploads the captured images to a master station. Furthermore, the acoustic modem can be used to transmit information such as the location of the AUV and to receive external information.

The area of an object to be inspected can be underwater fish farming pens, wherein the inspection comprises detecting if there is an unwanted hole or growth of unwanted organism in the underwater fish farming pens or if there is a presence of a large foreign body on or around the underwater fish farming pens.

According to a second aspect, the present invention relates to a method for underwater monitoring by means of using an autonomous underwater vehicle (AUV), the method comprising the steps of : moving the AUV in an area to be monitored, wherein underwater measurements of the AUV are performed and stored, locating the AUV with respect to a predetermined coordinate system, and associating coordinates of the AUV at any instant of time to the measurements at the same instant of time.

The measurements comprises at least one parameter from the group consisting of : water temperature, conductivity (salinity), dissolved oxygen, pH value of the water,

transmissivity of the. water, ammonia content and chlorophyll, wherein said at least one parameter is stored in a database along with the location coordinates of the AUV at the time of measurement.

The motion and the location of the AUV can be as described above and also the area to be inspected.

According to a third aspect, the present invention relates to a method for underwater surveillance and monitoring by means of using an autonomous underwater vehicle (AUV), the method comprising the steps of: moving the AUV in an area to be inspected, wherein -images of an object areas are frequently captured from the AUV and stored, underwater measurements of the AUV are performed and stored, locating the AUV with respect to a predetermined coordinate system, and associating coordinates of the AUV at any instant of time to the captured images and the underwater measurements at the same instant of time.

The location of the AUV, the measurement, the motion, the area to be inspected and the image capturing can be as described above.

According to a fourth aspect, the present invention relates to an autonomous underwater vehicle (AUV) system for underwater inspection and location, the AUV system comprising : o means for locating the AUV, o sonar to prevent an collision between the AUV and the surroundings,

o a receiver comprised in the AUV for receiving external sound waves and a transmitter for transmitting sound waves, o input means and output means, processing means and storage means, having stored therein at least one computer program, and o at least one capturing means for capturing images, wherein the captured images are stored in the storage means and wherein the coordinate at that instant of time is associated to the coordinate of the AUV system.

The location of the AUV, the motion, the area to be inspected and the image capturing can be as described above.

According to a fifth aspect, the present invention relates to an autonomous underwater. vehicle (AUV) system for underwater inspection and location, the AUV system comprising: o means for locating the AUV, o sonar to prevent an collision between the AUV and the surroundings, o a receiver comprised in the AUV for receiving external sound waves and a transmitter for transmitting sound waves, o input means and output means, processing means and storage means, having stored therein at least one computer program, and o at least one measuring means for measuring at least one of the following : water temperature, conductivity (salinity), dissolved oxygen, pH value of the water, transmissivty of the water, ammonia content and chlorophyll, wherein the measured quantities are stored in the storage

means and wherein the coordinate at that instant of time is associated to the coordinate of the AUV system.

The location of the AUV, the measurement, the motion, the area to be inspected and the image capturing can be as described above.

According to a sixth aspect, the present invention relates to an autonomous underwater vehicle (AUV) system for underwater inspection and location, the AUV system comprising: o means for locating the AUV, o a receiver comprised in the AUV for receiving external sound waves and a transmitter for transmitting sound waves, o a receiver comprised in the AUV for receiving external sound waves and a transmitter for transmitting sound waves, o input means and output means, processing means and storage means, having stored therein at least one computer program, o at least one capturing means for capturing images, wherein the captured images are stored in the storage means and wherein the coordinate at that instant of time is associated to the coordinate of the AUV system, and o at least one measuring means for measuring at least one of the following : water temperature, conductivity (salinity), dissolved oxygen, pH value of the water, transmissivity of the water, ammonia content and chlorophyll, wherein the measured quantities are stored in the storage means and wherein the coordinate at that instant of time is associated to the coordinate of the AUV system.

The location of the AUV, the measurement, the motion, the area to be inspected and the image capturing can be as described above.

Detailed description In the following the present invention, and in particular preferred embodiments thereof, will be described in greater details in connection with the. accompanying drawings in which, Figure 1 shows an example of a flowchart of the method, Figure 2 shows an example of a flowchart of the image processing method, Figure 3 shows an example of a flowchart of the state of alert, Figure 4 shows an example of a flowchart of the process of comparing measured values to set-point values, Figure 5 shows an example of an image of a mesh structure, Figure 6 shows an example of an image with background points removed, Figure 7 shows an example of detected holes, Figure 8 shows an example of detected large foreign bodies, and Figure 9 shows an example of detected growth and deposits.

The system consists of three or four main parts: 1. A master station, described in the text below.

2. A autonomous underwater vehicle AUV.

3. A long baseline system, described in the text below.

4. A docking station for the AUV, described in the text below.

The master station may be located on land or on a floating structure and consists of a computer (PC) with the necessary hardware devices including communication equipment. The computer runs software for setting the operating parameters of the system as a whole and the operating parameters of individual units of the system, optionally image processing software for processing the images accumulated by the AUV.

Figure 1 shows on embodiment of method. In the first step 1 the operator sets up the survey plan. The set up includes defining all the co-ordinates that the AUV will swim through in the specified order. The last co-ordinate must be the homebase, the docking station. Along with the position, instructions of turning on additional equipment and e. g. taking pictures or starting chemical measurements are needed. This is stored in a file called task file. The instructions can be strictly dependent on the co-ordinates or defined so that after a specific co-ordinate is reached the operations are performed at a fixed time interval. When setting up the survey the most important thing is to remember that this is an autonomous vehicle and one has to think about the limited energy and not make the survey. to long.

Next step 2 is to download the survey setup from the land-based PC to the computer on board the AUV. Here are two possibilities, downloading through a wireless local area network or to send the file by cellular or radio modem to the AUV. When the land based PC has downloaded the file the AUV starts its survey 3. The AUV starts each survey by calculating its position. For positioning in the area to be inspected, the AUV uses a long baseline system (LBL). Acoustic baseline stations serve as the reference points for positioning in long baseline systems. These stations are deployed near the corners of the underwater inspection site. Positions are measured with sub meter accuracy by interrogating the baseline stations and measuring the signal run time. Normally, a set of three stations is used to obtain high quality, unambiguous position fixes. The stations are usually mounted several meters off the sea floor, anchored by a line, and buoyed by their flotation collar. Rigid mounting on a pole is preferred in high surge applications and when centimeter accuracy is required. Baseline stations of this particular system support an automatic baseline survey (also known as sonar sing around). This eliminates the need to place stations precisely. The LBL system is turned on (from stand by to operating mode) by an acoustic signal from the AUV or the master station 4. The baseline stations reply by transmitting a signal, each station having its own special signal or key. The AUV listens and receives the signal. The AUV is capable of calculating the distance to each station using LBL methods and since their position is known the AUV is able to calculate its position.

The next step is to read the survey file to find out where to go i. e. the next co-ordinates.

This leads us to 5. If there are no more co-ordinates the AUV must be at its homebase 6,

because the homebase is defined as the last co-ordinates. The homebase is a docking station. This is an underwater unit that the AUV docks into and uses during data up-and downloading, battery recharging and during idle times. The AUV finds its way into the docking station by use of the LBL positioning system. The direction from which the AUV must swim in order to be able to enter the docking station under the correct angle is known a priori. The AUV swims towards the docking station from this angle and at the predetermined depth or height over bottom. If there are more co-ordinates the next one is read 7. The AUV now swims on 8 until it reaches the next co-ordinates. While the AUV swims on it continuously regulates its distance from the object (net), to be inspected, by using sonar or a digital camera. Also while swimming on the AUV checks for obstacles in front of it by using a sonar or a camera. With on-board image processing the AUV steers past the detected object. In 9 the AUV checks whether the position matches the co- ordinate. As defined by the operator in 1 the AUV reads from the task-file what to do at every position (co-ordinate) or time. This can include 10, to take pictures with e. g. side scan sonar, a video recorder or a camera, or 11 to start measuring chemical substances or'other parameters. The measurement can be one or more of the following parameters : temperature, conductivity (salinity), pH, transmissivity, dissolved oxygen, ammoniac content and chlorophyll. The picture or/and the measurement is stored in a database, 12 and 13, along with the co-ordinates and time. If the AUV is equipped with"on-board processing", the answer in 14 will be"yes". If not, the purpose of the survey is simply to gather data and therefore the next step is to check whether more co-ordinates exist 5. If the AUV has on-board processing installed, it starts the image processing 15, shown in Fig. 2, or/and the comparison of the measurements to predefined bounds of each parameter 16, shown in Fig. 4.

Referring now to Fig. 2 the image processing. This is a more detailed description of 15 in Fig. 1. It starts with reading the picture from the database 17.18 is the image processing. There are three objectives to the inspection 19. The first objective is to detect any holes in the mesh or net. The second objective is to estimate the extent of growth or other deposits on the structure. The third objective is to detect the presence of any large foreign bodies on or around the structure.

Images of sections of the mesh structure of the net are received from the imaging device for processing by an electronic circuit such as a digital computer. Each image is

assumed to be represented by intensity values sampled on the plane defined by the imager. In a pre-processing step each point in the image is classified based on observed image intensity as a foreground point or a background point. Background points are defined as any image points providing an unobstructed view beyond the mesh. All other image points are foreground points, regardless of whether they depict a portion of the mesh structure, growth, deposits, or a larger foreign body (see Figure 6).

The first objective is met by observing that any point in the plane defined by a section of mesh structure with maximum mesh dimension Dm can be at most a distance Dm/2 from the nearest point on the structure, In a front-parallel view of a section of mesh, this translates to a maximum observed mesh dimension dm = s*Dm in the image, and a maximum distance dm/2 of any image point from its nearest foreground point, where the scale factor s is determined by the focal length of the imager and the distance to the mesh. A hole larger than Dm is therefore signaled by the presence of an image point a distance larger than dm/2 away from its nearest foreground image point.

The detection of holes, as defined above, is effected by applying a dilation operator to the pre-processed image with a form element of diameter dm, and searching the resultant image for any residual background points. Following the dilation by an erosion operator with the same form element restores the approximate size and shape of the holes. The combined operation is the morphological closing transformation. Its effect is to only retain holes larger than the chosen form element (see Figures 5 through 7).

The second and third objective is met by observing that any point on a mesh structure of minimum width Ds is at most a distance D, 12 from the nearest point off the structure.

The presence of growth or debris on the structure which increases its apparent width, or the presence of a larger foreign body, is signaled by any image point a distance larger than ds/2 = s*Ds/2 from its nearest background point.

Any parts of the structure wider than Ds are detected by applying to the pre-processed image a morphological opening transformation with a form element of diameter ds, and searching the resulting image for any residual foreground points.

The third objective is met by applying the morphological opening transformation with a form element larger than the expected growth and deposits. This transformation retains only foreground regions larger than the chosen form element (see Figure 8).

The proportional area of the residual foreground points in the image provides an index of the extent of the accumulated growth and deposits. The areas detected as either holes or as large foreign bodies are removed from consideration prior to the summation of areas in fulfillment of the second objective (see Figure 9).

If nothing is detected the AUV keeps on swimming 20 and then continues checking for more co-ordinates as in 5 in Fig. 1. If the outcome is a state of alert 21 the AUV acts as shown in Fig. 3.

Referring to Fig. 3, the state of alert. This block diagram is a more detailed description of what happens in 21 in Fig. 2 and Fig. 4. The alert 22 process starts with 23 sending an acoustic signal to the master station through the docking station. This is done by an acoustic modem in the AUV and a corresponding acoustic receiver in the docking station. The message sent indicates the type of warning and the location of the cause.

The docking station alerts the human operator by sending the message to him by wire or through a wireless communication channel (cellular phone). If the message from the.

AUV is received successfully, it is acknowledged 24. If there is no acknowledge signal received the AUV tries again 25 until it has tried for a specified number of times, for example, 10. If no acknowledgement is received the AUV swims to home base (docking station) 26. Going home means to go to the last co-ordinate in the survey file. This last co-ordinate, as described before, should be the docking station. When the AUV is in the docking station it automatically uploads all the data to the land based station where the operator can visually examine the data and see whether something is wrong.

On the other hand if the AUV receives an answer it reads it 27 to see if this does affect the survey. In 28 it is asked whether to quit the survey or not. If it quits 26, it skips all but the last co-ordinates which are the co-ordinates of the docking station. If answer does not affect the survey the AUV keeps on swimming and checks the next co-ordinate. as in 5 in Fig. 1.

Fig. 4 describes the comparison of the predefined values of the environmental parameters to be measured to the actual measured values. This is a more detailed description of 16 in Fig. 1. It starts with 30 reading the measurements from the database and also the predefined bounds, both lower and upper. 31 is the start of the comparison process. In 32 the values are compared. If the measurements are within bounds the

AUV keeps on swimming 33 and as in Fig. 1 the next action is to 5 check to see whether there are any more co-ordinates remaining. If a parameter is outside the predefined bounds the state of alert 21 is initiated, as described in Fig. 3.