Method and system for quantifying a subject's exploration of a region of interest
The method calculates spatial exploration by measuring distances from a mesh of nodes within the region of interest, addressing the limitations of existing methods by providing an accurate and consistent score that reflects exploration thoroughness, independent of trajectory length and reducing manual annotation.
Patent Information
- Application Number
- PCT/IB2026/054135
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-05
- Filing Date
- 2026-04-27
- Publication Date
- 2026-11-12
AI Technical Summary
Existing methods for quantifying spatial exploration by a subject within a region of interest are flawed as they rely on trajectory length, arbitrary tile subdivision, or reference points, leading to inconsistent and subjective scoring that does not accurately reflect the thoroughness of exploration.
A method and system that calculates spatial exploration by determining the distance of the exploration trajectory from a mesh of nodes distributed throughout the region of interest, providing a score that is independent of trajectory length and capable of distinguishing central vs. peripheral exploration, without manual annotation.
The method generates a robust score that accurately reflects the thoroughness of exploration, is invariant under rigid transformations, and reduces the need for manual annotation, enabling efficient and consistent comparison across different scenarios.
Smart Images

Figure IB2026054135_12112026_PF_FP_ABST
Abstract
Description
TITLEMETHOD AND SYSTEM FOR QUANTIFYING A SUBJECT’S EXPLORATION OF A REGION OF INTEREST DESCRIPTIONField of the invention
[0001] The present invention relates to a method and a system for quantifying the spatial exploration of a region of interest by a subject, that is, for determining how thoroughly the region of interest or specific portions thereof are explored by a subject travelling along a trajectory within that region.
[0002] In this description, the term ‘subject’ refers to any moving subject, such as a person, an animal, a microorganism, but also a vehicle operated by a local or remote operator, a machine, or any other moving body.
[0003] Although the description generally refers to the case of a single subject, it is obvious that the invention can also be applied to determine the spatial exploration carried out by a plurality of subjects, whether considered individually or collectively, that is, by considering the exploration carried out by a set of subjects.
[0004] Moreover, by ‘spatial exploration’ an exploration is meant that takes place in a space of any dimensions, in particular, a three-dimensional or two-dimensional exploration.Prior art - Technical problem
[0005] There are many fields of technology and research in which it is important to estimate the exploration carried out in a given region of interest by one or more subjects or agents, who follow(s) trajectories within that region, in order to study the behaviour of those subject(s) under specific circumstances. A few non-exhaustive examples are mentioned below.
[0006] The subjects or agents can be:- animals, whose movements are tracked to study their behaviour during such activities as foraging, searching for a mate, seeking shelter,, etc., or during migrations, or during search-and-rescue operations involving, for instance, rescue dogs, or even to study how a particular experimentalcondition, such as the administration of a drug, influences the behaviour of such animals;human subjects, again to study the effects of clinical, pharmacological or other trials, in particular, involving people with debilitating conditions such as multiple sclerosis or Parkinson’s disease, or to assess whether object expected to be centres of attention are really able to draw the subjects’ attention in open or closed spaces, such as works of art in an exhibition, products on display in a retail outlet and the like, or to study attitudes towards elements of graphical interfaces such as website or social media screens, representing products, information and so on: in this case, the trajectories are described by the movements of the subject’s eyes, or by the displacements of pointing devices operated by the subject;- vehicles and mobile machinery in general, such as commercial vehicles for distributing products across a territory, drones or robots or vehicles in general for exploring potentially dangerous environments or scenarios of military interest, or for monitoring farms, factories, buildings in general, gardens, etc., and automated equipment for cleaning floors or mowing lawns.
[0007] Various methods are known for quantifying a spatial exploration performed by a subject. In some of those methods, such as the one described in Rodel et al. (2015), “Survival costs of fast exploration during juvenile life in a small mammal”, Behavioral Ecology and Sociobiology, 69, 205-217, but also the methods described in the references cited below, the quantification of a spatial exploration is based on the length of the trajectory, that is, on the linear distance travelled by the subject. However, this is an intrinsic characteristic of the trajectory only, i.e. a feature that does not take into account the physical space of exploration in which the trajectory is included. For example, according to such methods, a trajectory including a same path travelled a certain number of times by the subject will have an index that is higher the greater that number is, but a higher number of times does not mean a more thorough exploration of a region of interest including that trajectory.
[0008] Another method for quantifying spatial exploration, described in Beukeboom et al., (2024), “How to measure exploration: A combined estimation method”, Behavioural Processes, 216, 104998, involves truncating / approximat-ing the spatial coordinates, so as to create a tessellation of a given region ofinterest, i.e. a subdivision of this region of interest into a set of non-overlapping geometric shapes or tiles. The, the method quantifies a subject’s spatial exploration as a score that representing the percentage of tiles traversed by the trajectory the subject follows. This method has the disadvantage that the approximation of the spatial coordinates and, therefore, the number of tiles into which the region of interest is subdivided, is at the user’s discretion: this means that different exploration data are obtained according to how dense the subdivision is. Moreover, the method does not make it possible to distinguish the areas of the region of interest that are preferentially explored by the subject: for instance, the same score is obtained regardless of whether the trajectory preferentially crosses central or peripheral tiles of the region of interest, whereas it might be important to assign a higher score to one of these cases.
[0009] In other methods, described in Buard et al. (2011), “Visual exploration of large animal trajectories”, in 25thInternational Cartographic Conference (ICC’ 11 ). ICC (p.7) and in Petrini et al., (2016) “How vision and self-motion combine or compete during path reproduction changes with age”, Scientific reports, 6(1), 29163, the score quantifying the spatial exploration is based on calculating the distance of the trajectory from one or more specific reference points of the trajectory or the region of interest, which can include, for instance, the starting point of the trajectory and / or the centre of the region of interest. A disadvantage of this method is that a trajectory very distant from the reference point or points is assigned a high score, but this does not necessarily correspond to an accurate exploration of a region of interest that includes the trajectory.
[0010] Other methods, as described for instance in Hoch et al. (2020), “Where infants go: Real-time dynamics of locomotor exploration in crawling and walking infants”, Child Development, 91 (3), 1001-1020, involve filming the region of interest, manually annotating the subject’s exploratory behaviour frame by frame, and progressively assigning behaviour-related points to the subject, thus obtaining a final score. This is therefore a very labour-intensive method that is also characterised by excessive subjectivity in the scoring process. Baumann V. et al, “Towards a characterization of human spatial exploration behaviour”, Behavior research methods, 2025, pp. 57-65 provide a discussion of the most common movement measures currently used in human research on spatial exploration, and suggests new indices to describe the efficiency of exploration.Summary of the invention
[0011] It is therefore an object of the present invention to provide a method for quantifying the exploration of a region of interest by a subject travelling a trajectory within said region, i.e. to assign a score that indicates the accuracy by which the region of interest is explored while travelling said trajectory, which is free from the above -mentioned drawbacks of the prior art.
[0012] It is a particular object of the invention to provide such a method in which the score is independent of the length of the trajectory.
[0013] It is a further particular object of the invention to provide such a method in which the spatial exploration is given a score that is invariant under rigid transformations and uniform scaling of the region of interest and the exploration trajectory, so that scores calculated using different acquiring methods and units of measurement can be correctly compared.
[0014] It is a further particular object of the invention to provide such a method that assigns a higher score to trajectories which spend most of their time in the vicinity of the centre of the region of interest.
[0015] It is also a particular object of the invention to provide such a method that is more accurate in assigning scores than the prior art methods.
[0016] It is also a particular object of the invention to provide such a method not requiring excessive calculation, so as to be easily implemented in fast-executing and user-friendly software.
[0017] These and other objects are achieved by a method, as defined in independent claim 1 , for quantifying the spatial exploration of a region of interest carried out by a subject travelling an exploration trajectory. The objects are also achieved by a system for performing such quantification as defined in independent claim 20. Advantageous embodiments and modifications of the method and the system are defined in the claims depending on the above-mentioned independent claims, respectively.
[0018] According to one aspect of the invention, the method comprises the steps of:providing an electronic computer including a data input unit and a calculation unit;defining a system of spatial coordinates for the region of interest;delimiting the region of interest in said spatial coordinates;detecting the exploration trajectory by sampling, at consecutive time instants during the execution of the exploration trajectory, positions engaged by the subject in the region of interest;acquiring, by the calculation unit, data of the exploration trajectory in the form of position coordinates of said positions;acquiring, by the calculation unit, node coordinates of a number of nodes of a predetermined mesh of nodes lying within the region of interest; determining, by the calculation unit, the distance between each node of the mesh and the trajectory, i.e. the distance of each node from a position, among the positions of the trajectory, which is closest to that node, thus obtaining a plurality of raw node-trajectory distance values, i.e. a number of distance values equal to the number of nodes of the mesh, these distance values being calculated according to a same predetermined metric which can also differ from the standard 2-norm or Euclidean norm; calculating, by the calculation unit, a spatial exploration indicator starting from said plurality of raw node-trajectory distance values.
[0019] According to another aspect of the invention, the system for quantifying the spatial exploration of a region of interest carried out through an exploration trajectory by a subject comprises:a detector device configured to sample, at consecutive time instants during an execution time of the exploration trajectory, positions engaged by the subject in the region of interest while traveling along said exploration trajectory;an electronic computer including a data input unit and a calculation unit, wherein said calculation unit is configured to:delimit said region of interest by means of a system of spatial coordinates;acquire node coordinates of a number of nodes of a predetermined mesh of nodes lying within the region of interest;acquire data of the exploration trajectory in the form of position coordinates of the positions defined above;measure a distance of each node of the mesh from the position, among the positions defined above, that is closest to said each node, thus obtaining a plurality of raw node-trajectory distance values; calculate a spatial exploration index of the exploration trajectory starting from the plurality of raw node-trajectory distance values.
[0020] The term ‘mesh’ refers to a distribution of points called nodes that lie within the area of interest and are arranged either regularly, that is, spaced apart from one another by a uniform pitch, or even irregularly, provided that the position of each node is known in the chosen coordinate system.
[0021] The invention is therefore based on a novel approach to the problem of quantifying a spatial exploration, based on the calculation of the distance of the exploration trajectory from the nodes of a mesh distributed throughout the region of interest, which is different from the concepts of “trajectory length” and of “portion of the region of interest covered by the trajectory”.
[0022] In this way, thanks to the essential features of the method, namely, by defining a plurality of nodes lying within the region of interest and quantifying the spatial exploration on starting from the distances between each node of the mesh and the trajectory, significant effects are obtained that overcome the aforementioned disadvantages of the prior art.
[0023] Firstly, the proposed calculating method generates a result that is not directly influenced by the length of the trajectory. As illustrated below, if a score of the exploration is calculated according to the method of the invention and a same portion of a trajectory is travelled an increasing number of times, this does not result into an increase of the score.
[0024] Secondly, as the quantification of the spatial exploration, in the method of the invention, is based on the calculation of the distances between the trajectory and the nodes of a mesh that extends across the entire region of interest, instead of being based on the number of arbitrarily defined tiles that are traversed by the trajectory, a more robust score is obtained, i.e. a score that is less influenced by how the cells are drawn than in the prior art. Moreover, with the method according to the invention, it is possible to assign a higher score to the trajectories that are more central to the region of interest, since such trajectories are on average closer to the mesh nodes, compared to the score of more peripheral trajectories, which are on average further from the mesh nodes.
[0025] Thirdly, as the quantification of the spatial exploration is based, according to the invention, on the calculation of the distances of the trajectory from the nodes of a mesh that extends across the entire region of interest, instead of being based on the distances than from a single reference point, it is possible to one to avoid assigning a high score to a given trajectory simply because this trajectory predominantly traverse portions of the region of interest that are far from the point of origin, regardless of the actual exploration capacity of the trajectory.
[0026] Fourthly, the invention provides a method for quantifying a spatial exploration that makes it unnecessary to carry out tedious and time-consuming manual annotations of exploratory behaviours of agents from video files, frame by frame avoids tedious and time-consuming annotations made directly on video files.
[0027] In a first preferred embodiment of the method, and in a corresponding first preferred embodiment of the system, the step of calculating a spatial exploration indicator, for which the calculation unit is configured, comprises a step of calculating an exploration score in the form of a raw spatial exploration score of the exploration trajectory as an average of the raw node-to-trajectory distance values. This score is a single number that summarises the effectiveness of the exploration.
[0028] In a second preferred embodiment of the method, and in a corresponding second preferred embodiment of the system, the step of calculating of a spatial exploration indicator, for which the calculation unit is configured, comprises, after the step of determining the distance of each node from the exploration trajectory, a step ofgenerating a heatmap in the selected system of spatial coordinates, by the calculation unit and starting from the plurality of raw node-trajectory distance values determined as mentioned above. A heatmap, otherwise known as a heat map or as intensity map or diagram, refers to a diagram plotted in the chosen system of spatial coordinates corresponding to the region of interest and comprising, at the mesh nodes, respective coloured areas, each having a colorimetric feature or a density of symbol repetition monotonically depending on the respective distance values. This way, animmediate indication is obtained of the areas lying within the region of interest where the exploration has been most successful.
[0029] The step of delimiting said region of interest can comprise a step in which the calculation unit acquires intervals of said spatial coordinates representing said region of interest.
[0030] In respective modifications of the embodiments of the method that have been defined above, in the case of a region of interest having a polygonal shape with a given number of vertices, for example in the case of a rectangular region of interest, the following steps can be provided:inputting the vertex coordinates of said number of vertices of the polygonal shape into the data input unit of the electronic computer;transferring the vertex coordinates from the data input unit to the computer’s calculation unit, andthe step of acquiring the spatial coordinate intervals representing the region of interest by the computer’s calculation unit includes a step of calculating the spatial coordinate intervals representing the region of interest, performed by the computer’s calculation unit, and starting from the vertex coordinates. In other words, an automatic procedure is provided for delimiting or defining the polygonal region of interest, based solely on the vertex coordinates.
[0031] In a corresponding modification of the two embodiments of the system defined above, the data input unit is configured to receive and transfer to the calculation unit the coordinates of the vertices of the polygonal shape, and the calculation unit is configured to calculate the spatial coordinate intervals representing the region of interest based on the vertex coordinates.
[0032] As an alternative to the above, the region of interest can be of any shape, including a polygonal shape, and the following steps can be included:inputting the spatial coordinate intervals representing the region of interest into the data input unit of the electronic computer;transferring the spatial coordinate intervals representing the region of interest from the data input unit to the computer’s calculation unit, thus causing the step of acquiring the spatial coordinate intervals to be performed by the calculation unit.In other words, in this case the region of interest is defined manually by an operator. For instance, the input form of the spatial coordinates of the region ofinterest into the data input unit can be a curve or a plurality of curves analytically defined according in the selected spatial coordinate system and representing the outline of the region of interest.
[0033] Even if the drawings and the examples always refer to trajectories lying in a plane and to planar regions of interest, this is not a limitation of the method and of the system of the invention. Indeed, the drawings and the examples can be readily modified by a person skilled in the art to encompass the case of three-dimensional trajectories, for example to the case of trajectories lying on such three-dimensional surfaces as quadric surfaces. In this case, accordingly, the region of interest is three-dimensional as well, for example it can be a space portion defined by such a three-dimensional surface, or a surface portion of said three-dimensional surfaces.
[0034] Moreover, even if the drawings and examples always refer to orthogonal Cartesian coordinate systems, this is not a limitation of the method and of the system according to the invention. Indeed, the drawings and the examples can be readily modified by a person skilled in the art to encompass the case of coordinate systems of a different type, such as a two-dimensional polar coordinate system or a three-dimensional spherical or cylindrical coordinate system, which the person skilled in the art will be able to most appropriately select based on the shape of the trajectory and / or the region of interest.
[0035] In respective advantageous modifications of the embodiments defined above, the step of acquiring the node coordinates includes a step of calculating the node coordinates by the calculation unit of the electronic computer which, in corresponding modifications of the system, is configured to calculate said node coordinates. In this case, the mesh is therefore automatically defined by the calculation unit, preferably in a manner depending on the structure of the exploration trajectory, as illustrated below, so as to achieve a favourable balance between the accuracy of the method and the required computational power, both of which increase as the number or density of the nodes of the mesh increases.
[0036] In particular, the mesh of nodes can be a matrix of nodes spaced apart from one another by a predetermined pitch, i.e. a matrix in which each node is at a same distance from each of its neighbouring nodes. In other words, the nodes have a uniform distribution with a predetermined pitch. In a corresponding modification of the system, the calculation unit is configured to define the valueof the pitch between the nodes of the mesh and to consequently calculate the node coordinates.
[0037] In this case, advantageously, before the step of acquiring the node coordinates via calculation by the calculation unit, the above-mentioned steps are performed of detecting the exploration trajectory and of acquiring the data of the exploration trajectory, in addition to the steps of:calculating partial lengths of consecutive portions of the exploration trajectory between consecutive positions sampled along the exploration trajectory during the execution time of the exploration trajectory, thus obtaining a partial length distribution;extracting a statistic of said partial length distribution, selected between the maximum value and the adjacent upper value or upper fence of said distribution;furthermore, in the step of calculating the node coordinates by the calculation unit, the pitch is selected not exceeding said statistic of the partial length distribution. As a matter of fact, the inventors have observed that, for a given trajectory in a given region of interest, as the pitch decreases, the score converges to a stable value, substantially reaching that stable value when the pitch becomes equal to such a statistic, in particular, when the pitch becomes equal to the upper fence of the partial length distribution. In other words, by selecting a pitch below this statistic, the score remains essentially unchanged. The accuracy of the method, for a given trajectory in a given region of interest, increases by decreasing the pitch down to the statistic of the partial length distribution, i.e. by increasing the node density up to a value corresponding to that statistic. However, no substantial gains in terms of accuracy are obtained by further decreasing the pitch below the statistic of the partial length distribution, i.e. by increasing the node density beyond a certain value.
[0038] Choosing the statistic as the upper fence is advantageous as it allows any isolated, insignificant maximum values of partial lengths to be excluded from calculation. The statistic of the partial length distribution may, however, also be any percentile or quartile.
[0039] For the reason stated above, it is preferable that the pitch between mesh nodes be chosen to be substantially equal to the statistic of the partiallength distribution, so as to achieve the maximum precision of the method without unnecessarily increasing the computational burden of the method.
[0040] Accordingly, the system’s computer calculation unit is configured to sample consecutive positions along the exploration trajectory, calculate the lengths of the consecutive portions between them, extract the aforementioned statistic, and calculate the node coordinates by setting the pitch to a value not exceeding the statistic of the partial length distribution, in particolare, to a value substantially equal to that statistic.
[0041] As an alternative, the steps are provided of:Imputing the node coordinates into the data input unit;transferring the node coordinates from the data input unit to the calculation unit, so as to cause the step of acquiring said number of said node coordinates to be performed by the calculation unit.In other words, in this case, the node mesh can be defined by a user of the method by entering the node coordinates, however they are distributed in the region of interest, rather than through a routine performed by the computer’s calculation unit. This can be advantageous if computation time is not a determining factor, and if the user wishes to increase the degree of precision of the exploration indicator, which can be useful when one wishes to distinguish between similar exploration trajectories with greater precision.
[0042] The exploration trajectory can be a trajectory physically travelled by the subject lying within the region of interest during the execution time of the trajectory.
[0043] In this case, the subject or the subjects can be, for instance, one or more people who are physically present in the region of interest and who receive stimuli to perform movements, i.e. trajectories. In particular, as mentioned above, applications in the medical field are possible, such as the study of the effects of clinical, pharmacological or other types of trials, by quantifying the ability of subjects with pathologies and / or disabling conditions — such as multiple sclerosis, Parkinson’s disease, or visual or sensory impairments in general — to explore a region of interest. Further similar applications can concern the assessment of the relationship between the exploratory capacity of an environment and the neuro-cognitive and motor development of children. The method can also be used in a commercial context, in order to assess the attractiveness exertedon a population by such focal points as goods displayed in a retail outlet or the like. In similar applications, the points of attraction can be works of art displayed in a museum or exhibition, to assess the ability a work of art to attract the general public.
[0044] The subjects can also be animals, in ethological applications, where the movements of individual specimens or populations are tracked to study their behaviour during activities such as foraging, searching for a mate for potential mating, seeking shelter and so on, or during mass movements such as migrations, transhumance, farm operations, etc. Applications are also possible in the veterinary field, similar to the medical applications with human subjects mentioned above, or in pharmaceutical research for the study of the collective behaviour of populations of laboratory animals. A particular application is possible in the training / selection of animals such as rescue dogs, where it is important to assess their ability to explore an open or enclosed environment.
[0045] The subjects can also be microorganisms, enabling a function of quantify spatial exploration in microorganism-tracking devices incorporated in microscopes.
[0046] The subject can also be a vehicle or, more in general, a moving equipment. In particular, as mentioned above, applications are possible in the domestic sphere, in order to quantify the ability to explore a region of interest, such as a floor, by robotic household appliances like vacuum cleaners and floor cleaners, or a lawn, by a robotic lawnmower. Applications are also possible in the commercial sector, to assess the exploration efficiency of geographical areas by vehicles used for shipping and, in particular, for selling / delivering various types of goods. Other possible vehicular applications relate to the field of territorial surveillance for scientific research, civil protection or military purposes, to assess the ability to explore an inaccessible and / or dangerous area in case of unfavourable events such as landslides, mudslides, the presence of explosive devices such as mines, enemy ambushes and so on, by such vehicles as drones, as well as other types of robots, and vehicles in general including boats, in particular, in the military sector. Applications are also possible in the field of security of civil and / or industrial installations or private and public areas, such as farms, industrial plants, gardens, parks, sport areas, public squares and urban areas in general, to assess the ability to explore such installations and areas by robotsand vehicles of various types including, in particular, drones or different driverless vehicles.
[0047] The detection of a trajectory can be performed by various types of tracking systems, without particular limitations, for instance, motion capture systems with or without markers, such as NEXUS Vicon, video cameras, VR motion tracking, and eye tracers, as illustrated below.
[0048] For example, when the exploration is performed that are by subjects physically present in the region of interest during the execution of the trajectory, the step of detecting the exploration trajectory can be carried out by a device selected from the group comprised of:a device comprising a plurality of video cameras, each having a field of view covering a respective portion of the region of interest;a device comprising a single video camera having a field of view covering the entire region of interest;a microorganism tracking device incorporated in a microscope;a device comprising a tracer of a predetermined sensor, the tracer configured to be worn / installed on a human or animal, or on moving equipment respectively, and having a reception field including the region of interest.
[0049] In other cases, the region of interest can be a portion of a scenario observed by the subject, and the exploration trajectory is drawn onto the region of interest by an optical axis oriented by the subject through his / her own eye movements. In such cases, the step of detecting the exploration trajectory can be performed by an eye-tracking device.
[0050] In yet other cases, the region of interest can be a portion of a computer screen or a similar device, controlled by the subject via pointing devices, such as a mouse or similar device. In these cases, the step of detecting the exploration trajectory can be performed by a cursor tracking device for tracking a position on the screen of a cursor associated with the pointer means.
[0051] The method according to the first embodiment defined above, in which a score is calculated as an indicator of spatial exploration, can further include a step of normalising the raw spatial exploration score so as to obtain a normalised spatial exploration score set between a minimum score value and a maximum score value. In particular, the minimum and maximum score values can be assigned as 0 and 1 , respectively. This allows for the comparison of differentexploration scenarios, provided that the region of interest is the same and that the same minimum and maximum values for the normalised score are chosen for all scenarios. Preferably, the minimum and maximum score values are defined such that they correspond to the subject’s minimum and maximum exploration conditions of the region of interest, respectively, so as to facilitate the immediate interpretation of the results of the analyses performed using the method.
[0052] Advantageously, the spatial exploration score is calculated as a normalised spatial exploration score using a formula chosen fromandwhere:S is the normalised spatial exploration score;S is the raw spatial exploration score, calculated as the arithmetic mean of all raw node-trajectory distance values n, i=1...N;Smaxis the maximum raw spatial exploration score, i.e. the score of the trajectory with minimum exploration for the given region of interest, calculated by the steps of:for each vertex of said vertices (Vh, h=1 ...H) of said polygonal shape of said region of interest:calculating of a distance from each of said vertices to each node of said number of nodes of said mesh, thus obtaining a plurality of vertex-node distances;calculating of an average vertex-node distance between each of said vertices and said nodes,thus obtaining a plurality of average vertex-node distances; defining said maximum raw spatial exploration scoresmaxas the maximum value among said average vertex-node distances.Smaxcorresponds to the raw spatial exploration score of a trajectory of a subject who remains stationary at the vertex furthest from the mesh, on average. Forsome polygons, such as rectangles and squares, Smaxcan be calculated analytically, based on the lengths of the sides of the polygon.
[0053] By formula [I], a normalised spatial exploration score is obtained in a manner easily implementable by a person skilled in the art, with the above-mentioned advantages of comparability between different situations and intuitiveness. Formula [II], on the other hand, provides score values that are better distributed across the entire interval [0,1].
[0054] In corresponding embodiments of the system, the calculation unit is configured to normalise said raw spatial exploration score in the various ways indicated above.Brief description of the drawings
[0055] The invention will now be illustrated by describing certain embodiments, by way of example and not by way of limitation, with reference to the accompanying drawings, in which:- Fig. 1 schematically shows the system according to the invention;Fig. 2 is a diagram showing the variation over time of the position coordinates engaged by a subject along an exploration trajectory;Fig. 3 shows in greater detail the region of interest and the exploration trajectory of Fig. 1 ;- Fig. 4 is a flowchart representing the method according to the invention, in two basic embodiments;Fig. 5 is a flowchart of the step of delimiting the region of interest, according to two possible modifications of the embodiments shown in Fig. 4;- Fig. 6 is a flowchart that refers to an embodiment of the method in which specific preparatory steps are provided before the step of acquiring data of the exploration trajectory;Fig. 7 shows schematically the region of interest and the exploration trajectory as in Fig. 3, along with the parameters for calculating the pitch of node mesh, according to a modification of the method;Fig. 8 schematically shows a portion of the exploration trajectory of Fig. 7;Fig. 9 is a flowchart that refers to an embodiment of the method alternative to the embodiment of Fig. 6, still in connection to the step of acquiring data of the exploration trajectory;Fig. 10 schematically shows the system according to an embodiment of the invention, in which an optical scanning device is provided comprising a plurality of video cameras, each having a field of view including a respective part of the region of interest;Fig. 11 schematically shows the system according to another embodiment of the invention providing a sensor for tracking a subject under examination, the sensor having a reception field including the entire region of interest;Fig. 12 schematically shows the system according to a further embodiment of the invention, provided with a sensor configured to detect the position of a subject’s optical axis of in order to identify the points of a trajectory drawn by said optical axis on a scenario observed by the subject;Figs. 13a, 14a and 15a show the exploration trajectories in respective examples of application of the method;Figs. 13b, 14b and 15b are diagrams showing how the exploration score changes during the exploration according to the method of the invention and according to respective prior art methods for quantifying spatial exploration;Fig. 16 shows a non-parametric paired t-test for the statistical comparison of the results of a test on multiple simultaneous trajectories.Description of preferred embodiments
[0056] With reference to the accompanying drawings, a method and a system 10 for quantifying the spatial exploration of a region of interest 1 that a subject 3 performs by travelling along an exploration trajectory y are described below.
[0057] As schematically shown in Fig. 1 , system 10 comprises a detector device 20 configured to sample, at consecutive time instants tk, k=1 ...M, of an execution time T of exploration trajectory y by subject 3, positions Gk= G(tk), k=1 ...M engaged by subject 3 in region of interest 1 while traversing exploration trajectory y, Figs. 2 and 3. System 10 further comprises an electronic computer30 including a data input unit 31 and a calculation unit 32 arranged to receive data from data input unit 31.
[0058] As shown in Fig. 4, the method comprises, firstly, preliminary steps S10 of prearranging electronic computer 30, and S20 of defining a system of spatial coordinates in region of interest 1. In the example shown in Figs. 1 and 3, a two-dimensional orthogonal Cartesian coordinate system x,y is shown on a region of interest 1 extending over a plane IT. However, upon reading this description, a person skilled in the art will be able to implement the method also on regions of interest extending, for example, over curved surfaces and / or in the three-dimensional space, conveniently selecting an appropriate Cartesian, cylindrical, spherical or other coordinate system depending on the spatial configuration of region of interest 1 , for example a latitude / longitude system if region of interest 1 is an extensive georeferenced region.
[0059] Method operational steps then follow, in a sequence not necessarily in the same sequence as indicated below, unless otherwise stated.
[0060] As further indicated in Fig. 4 and shown in greater detail in Fig. 5, during step S30 of delimiting or defining region of interest 1 , calculation unit 32 of electronic computer 30 acquires, in step S35, an analytical representation of region of interest 1 in terms of (x,y) coordinates of the system of spatial coordinates. This analytical representation consists of a geometric shape, in this case a plane geometric shape in (x,y) coordinate system, analytically defined as the area inside a contour curve expressed by an equation of the form <p(t)= (x(t),y(t)) where t is a parameter of the curve, or as a set of intervals Ax, Ay [yf,yj”], j=1 ...J of one y-coordinate for values in corresponding intervals Ax, Ay Axj = [xj-i ,xj ] of the other coordinate, or even two-dimensional intervals Ax, Ay. Without loss of generality, reference will be generally made to the acquisition of Ax, Ay intervals representing region of interest 1.
[0061] As Figs. 1 and 3 further illustrate, a mesh 2 of N nodes Pi, i=1 ...N is defined within region of interest 1 , distributed throughout region of interest 1. The specific features of the invention are based on these nodes, as explained in greater detail below.
[0062] Step S30, which involves defining region of interest 1 (Fig. 4), can be carried out in various ways. In one case, as shown in Fig. 5, the analytical representation of region of interest 1 , for example in one of the forms indicatedabove, can be directly entered by an operator, step S31 , into data input unit 31 of electronic computer 30, and then simply transferred to calculation unit 32, step S32, which thus acquires it, step S35. Alternatively, as Fig. 5 further illustrates, in the case of polygonal regions of interest such as rectangular region of interest 1 shown in Fig. 4, coordinates Xh,yh, of vertices Vh, h=1 ...H, are entered into data input unit 31 by an operator, step S33, and then transferred, step S34, to calculation unit 32, and step S35, in which calculation unit 32 acquires the analytical representation of region of interest 1 , is carried out by calculating, in step S36, the polygonal shape enclosing region of interest 1 based on vertex coordinates Xh,yh, h=1 ...H.
[0063] With further reference to Figs. 1 and 4, the method also comprises a step S40 for detecting exploration trajectory y by means of detector device 20 of system 10. Detection step S40 comprises a step S45 of sampling positions Gk=G(tk), k=1 ...M engaged by subject 3 in region of interest 1 , at a plurality M of respective consecutive time instants tk, k=1 ...M during execution time T=tM-ti in which exploration trajectory y is travelled by subject 3.
[0064] Concerning region of interest 1 and how the subject interacts with it, three basic configurations can be distinguished.
[0065] In the first configuration, subject 3 interacts with region of interest 1 by being physically located within it and physically travels exploration trajectory y. In this case, subject 3 can be a person, an animal or a moving equipment, for example a motor vehicle, an airplane, a drone, a boat, a robot such as a robotic vacuum cleaner or floor cleaner, or a robotic lawnmower.
[0066] In some modifications of this configuration, while subject 3 travels along exploration trajectory y, a detector device ‘frames’ region of interest 1 and detects the position of subject 3 within it. Subject 3 can optionally be equipped with a tracer element 8 to which the selected sensor is sensitive.
[0067] For example, step S40 of detecting exploration trajectory y travelled by subject 3 can be performed by means of an optical detector device 20 comprising a plurality of video cameras 6q, q=1...Q, as in Fig. 10, each arranged to frame a respective part of region of interest 1 , or an optical detector device 20 comprising a single camera 6, as in Fig. 1 , having a field of view 4 that includes entire region of interest 1.
[0068] This configuration also includes the case, not shown, in which the subjects are microorganisms, and the detector device or tracking device is configured for detecting said microorganisms and is incorporated in a microscope.
[0069] In certain modifications of this configuration, compatible with modifications defined above, system 10 or detector device 20 can comprise a specific tracer 8 configured to be worn or installed on subject 3 or on each subject 3, such as a human, an animal or a moving equipment. In particular, as shown in Fig. 11 , tracer or tracers 8 can be associated with a sensor 7, which can be, but is not limited to, an optical sensor, configured to detect the position of tracer or tracers 8, where sensor 7 has a reception field 5 that includes region of interest 1. In some alternatives, not shown, sensor 7 is a localisation sensor configured to calculate its own position using a global navigation satellite system (GNSS), such as GPS, Galileo or others, or to calculate the position using a local reference system, for example in the case where sensor 7 is configured to interact with transmitters available in the vicinity of region of interest 1.
[0070] Examples of combinations of sensors 7 and tracers 8 can include, within the field of optical tracking, retroreflective markers used for motion capture, VR headsets or other devices worn by subject 3 themself, which allow the position to be tracked within reference region 1 , such as a room or an outdoor area.
[0071] In a second configuration, subject 3 interacts remotely with a real region of interest 1 by his / her own senses, typically sight, or via a wearable sensory exploration device equipped with a sensing component such as a camera, following exploration trajectory y by an own natural optical axis 9, or by an optical axis generated by the wearable device. For instance, region of interest 1 can be a portion of a scenario observed by subject 3, as shown in Fig. 12, or can be a portion of a screen such as a monitor or an equivalent device, and exploration trajectory y is drawn on region of interest 1 by an optical axis 9 which is oriented by subject 3 by his / her own eye movements. In this case, step S40 of detecting exploration trajectory y can be performed by means of an eye-tracking device 40 for tracking the eye movements of subject 3, and exploration trajectory y can be displayed on screen 35 of a computer or similar device, which can be, for instance, electronic computer 30 used for performing the method of the invention, as shown, for instance, in Figs. 1 and 3.
[0072] The tracking device for tracking optical axis 9, or the eye-tracer, can be of a fixed type, comprising, for instance, one or more video cameras observing the subject, or can be of a mobile type, wearable by the subject themself, for example, it can comprise special glasses equipped with video cameras.
[0073] In a third configuration, not shown, the subject interacts via a pointing device with a virtual region of interest 1 such as a screen 34 generated to the monitor of a computer, or similar device, by a locally resident software or a website. This device can comprise, for instance, electronic computer 30 used for executing the method as shown in Figs. 1 and 3. In this configuration, the detector device is integrated in the computer, and can comprise a software configured to record the movements and halts performed by a cursor controlled by the pointing device on screen 34 at virtual points of interest within region of interest 1.
[0074] Referring once again to Figs. 1 and 4, the method also comprises a step S50 in which calculation unit 32 acquires data of exploration trajectory y. These data can take the form of position coordinates x(tk),y(tk), k=1 ...M of sampling positions Gk=G(tk), k=1 ...M along exploration trajectory y engaged by subject 3 at times tk of execution time T. This example of a two-dimensional trajectory lying on plane IT, represented by Cartesian coordinate system (x,y), can be easily extended to a different coordinate system lying on plane IT, or to a coordinate system suitable for describing a three-dimensional curve, for example, a curve that lies on a surface defined analytically in space, such as a quadric surface.
[0075] Still with reference to Figs. 1 and 4, the method also comprises a step S60 in which calculation unit 32 acquires node coordinates (xi,yi) of nodes Pi(xi,yi), i= 1 ...N of mesh 2 of nodes lying within region of interest 1. Preferably, in this step, calculation unit 32 does not merely receive node coordinates (xi,yi), i=1 ...N from an operator, for example via data input unit 31. On the contrary, step S60 of acquiring node coordinates (xi,yi), i=1 ...N advantageously comprises a step S65 of automatically calculating node coordinates (xi,yi), i=1 ...N by calculation unit 32. In other words, the method imposes, regardless of the operator, mesh 2 of nodes onto region of interest 1 , for instance, in the form of a matrix in which nodes Pi, i=1 ...N have a predetermined pitch p with respect toone another, i.e. they are uniformly spaced apart from one another throughout mesh 2, as shown in Figs. 1 and 3.
[0076] In the present method, the selection of pitch p occurs in a manner that takes into account how exploration trajectory y unfolds within region of interest 1. To this purpose, steps S40 and S50 of detecting exploration trajectory y and acquiring data of exploration trajectory y, are performed before step S60 of acquiring node coordinates (xi,yi), when this acquisition is carried out by calculating the coordinates (xi,yi) of nodes Pi, i=1 ...N by calculation unit 32, step S65.
[0077] For example, as shown in Figs. 6-8, in order to select, in step S59, a pitch p of mesh 2 that takes into account the structure of exploration trajectory y, calculation unit 32, based on the data of exploration trajectory y acquired in step S50, performs step S57 of calculating partial lengths Zk, k=1 ...M of consecutive portions yk k=1...M lying between consecutive positions Gk,Gk+i k=1...M sampled at consecutive time instants tk,tk+i k=1 ...M along exploration trajectory y, and further performs a step S58 of extracting a statistic of the partial length distribution, preferably the upper adjacent value or upper fence of that distribution, or even the maximum value of partial lengths Zk, k=1...M calculated as mentioned above. As well known, the upper adjacent value can be calculated as Q3 + 1.5 IQR, where IQR= Q3-Q1 , Q1 is the first quartile and Q3 is the third quartile.
[0078] Subsequently, calculation unit 32 performs step S59 of selecting pitch p as a value not exceeding the statistic, and even more preferably as a value substantially equal to the statistic of the distribution of partial length Zk, k=1 ...M, and is therefore able to perform step S60 of acquiring coordinates (xi,yi) of nodes Pi by calculation step S65.
[0079] The statistic of the partial length distribution can even be any percen-tile / quartile.
[0080] However, as shown in Fig. 9, it is possible to manually assign pitch p, i.e. to manually define mesh 2 of nodes Pi, so that it takes into account the structure of exploration trajectory y, through a step S61 of inputting node coordinates (xi,yi), i=1 ...N into data input unit 31 by the operator, and through step S62 of transferring these coordinates to calculation unit 32, so that step S60 of acquiring node coordinates (xi,yi) is carried out by inputting and transferring instead of calculating them by calculation unit 32, step S65.
[0081] As an alternative to the methods described above, pitch p can be defined through an iterative algorithm, which takes into account how the exploration index changes as mesh 2i, 22, 2a... becomes finer, the index indicating the iteration number of the iterative algorithm. More specifically, in the first iteration, a relatively coarse mesh 2i having a pitch pi is considered, and an exploration score Si of the trajectory y is calculated accordingly. In subsequent iterations k, increasingly fine meshes 2k, i.e., progressively smaller pitches pk are considered, and further exploration scores Sk are calculated accordingly, which does not require recalculating distance of all the nodes Pi from mesh 2k from exploration trajectory y, but only the distance of the nodes added in current iteration k, since the distances of the nodes of mesh 2k-i have already been calculated during previous iteration k-1. At each iteration, the quantity |Sk-i -Sk | is also calculated, that is, the difference between the score of current iteration k and the score of previous iteration k-1. The iterations are discontinued when this difference falls below a predetermined value small enough to be considered negligible, since further refinement of mesh 2k would not yield any improvement of the precision of the exploration score, and the score calculated in current iteration k is assumed as the exploration score.
[0082] The use of such an iterative technique is advantageous, for example, when one wishes to calculate the score to within a given tolerance, or when the trajectory sampling is very fine. As a matter of fact, in the latter case, the use of a direct calculation technique would result in a very dense mesh 2, and would therefore entail a high computational cost.
[0083] Once mesh 2 has been defined, calculation unit 32 performs a step S70 of determining the distance n, in a metric that may be the Euclidean metric or a different metric, between each node Pi of mesh 2 and a particular position Gk,i that is closest to that node Pi, among positions Gk along exploration trajectory y. This yields a set of raw node-trajectory distance values n, i=1 ...N, each corresponding to a respective node Pi. Step S70 of determining these distances n can also be performed using an optimisation technique based on Kd-tree structures, which is known to those skilled in the art.
[0084] The distance function can be that defined by any norm p. For example, when p=2, the standard Euclidean distance is obtained. As an alternative, the distance function can be defined by the user depending on the specific caseunder consideration. For example, if the region of interest is a portion of the Earth’s globe, it is preferable to use the geodetic distance, or an approximation thereof, if the projection is unknown. The geodetic distance can also be advantageously used even if the area of interest is a concave polygon. In this case, one can use well-known iterative algorithms, also known as ‘greedy’ algorithms, or even algorithms based on the solution of partial differential equations. As an alternative, it is possible to use an anisotropic metric, i.e. a metric that ‘favours’ one direction over another, and which yields a distance function whose isolines are ellipses rather than circles, in the case of a point source. Similarly, rather than favouring a given direction of expansion, a user of the method might be interested in giving higher or lower importance to the proximity to a given set of points. In this case, a graph can be defined within the mesh that weights the edges differently depending on the nodes they connect. Once this graph has been constructed, the calculation of the distances can be performed using algorithms known to those skilled in the art, for example Dijkstra’s algorithm.
[0085] Referring again to Fig. 4, in the method according to the invention, step S70 of determining raw node-trajectory distance values n, i=1 ...N, forms the basis for calculating a spatial exploration indicator, steps S80 and S90.
[0086] In one embodiment, the spatial exploration indicator is a spatial exploration score or index; in particular, a raw spatial exploration score S of the exploration trajectory y based on raw node-trajectory distance values n,i=1 ...N. This raw spatial exploration score S is chosen as an average of the raw node-path distance values n, i=1 ...N, specifically equal to the arithmetic mean of these raw values:
[0087] Preferably, however, a step S85 of normalizing the raw space exploration score S is also provided, shown by dotted line in Fig. 4 as an optional step. This is a step of calculating a normalised space exploration score S set between a minimum value and a maximum value, these values being preferably equal to 0 and 1 , respectively. The way to calculate normalised spatial exploration score S from raw score S is preferably selected so that the minimum value and the maximum value correspond to conditions of minimum and maximum exploration, respectively, the region of interest 1 by subject 3, i.e. so that the normalisedspatial exploration score S decreases as the raw score S increases. This can be obtained in a simple way, for example, by the formula:or by the formulawhere S is the normalised spatial exploration score, S is the raw spatial exploration score, calculated as the arithmetic mean of all raw node-trajectory distance values n, i=1...N, and Smaxis the maximum raw spatial exploration score, i.e. the score of the trajectory with minimum exploration of the region of interest, obtained by calculating, for each vertex of the polygon defining the region of interest, the distances of that vertex from the nodes of the mesh, calculating the average value of these distances, and taking the maximum average vale among the average values corresponding to each vertex.
[0088] In another embodiment of the method, the spatial exploration indicator is a heatmap in spatial coordinate system (x,y) selected for representing region of interest 1. This heatmap is calculated by calculation unit 32, step S90, as shown again in Fig. 4. Heatmap comprises areas at nodes Pi, i=1 ...N of mesh 2, in each of which a predetermined colorimetric feature has an intensity univocally depending on the value of node-trajectory distance n of the corresponding node Pi. The colorimetric feature can be the red or blue component, or other colours, in a colour space such as RGB, or the brightness, or the saturation and so on, with monochromatic heatmaps also being possible. A similar effect can be achieved, for example, by using a local density of repetition of a symbol such as a dot, a cross and so on, instead of a true colorimetric variable.
[0089] In a further embodiment, not shown in the figures, the spatial exploration indicator can be the table or matrix of the values of node-trajectory distance n, or of values related thereto that vary inversely with respect to the node-trajectory distance.
[0090] The user of the method and / or system can also define a sub-region of interest within the entire region of interest 1 , in which the subject moves, calculating the exploration score in one of the ways described above. In other words, for example, the exploration score is calculated in the same way as explainedabove, creating mesh 2 within the sub-region of interest (not shown), but normalising the score using formula [I] or formula [II] with the value of Smaxcalculated across entire region of interest 1. In this way, the user benefits from obtaining more precise information within more specific sub-regions. For example, the area where the exploration indicator is calculated, such as a portion of a room where attractors or focal points are present, may be smaller than the entire region where the subject moves, such as the entire room.
[0091] Figs. 13a-15b relates to three examples of calculation of an exploration indicator with reference to three different trajectories y’, y” and y”’ within a region of interest 1. These examples make it possible to compare the method of the invention with the methods currently used to measure spatial exploration that, as already mentioned, are based on the length of the trajectory (Figs.13a / b), on the subject’s distance from a fixed point (Figs. 14a / b) and on the area covered by the subject while travelling along the exploration trajectory (Figs.15a / b).
[0092] Fig. 13a shows an exploration trajectory y’ that extends from the centre to the periphery of a region of interest 1 , and that consists of five round trips along a segment 11 , starting from one of its endpoints 14a and 14b. The execution time T of exploration trajectory y’ is 1000 time units, for example 1000 s. The diagram in Fig. 13b shows the values of the distance travelled by the subject and the score calculated over time according to the method of the invention and to the prior art, more precisely, by sampling k instants (ti ...tk) and using the method of the invention I the prior art to calculate spatial exploration for each instant tk. It can be observed that, as the exploration proceeds, the score based on the length of the trajectory, according to the prior art, increases monotonically, in a substantially linear manner, indicating a progressively greater exploration over time, whereas the score according to the method of the invention, in a manner that is more consistent with the actual exploration of the region of interest 1 , increases up to a point in the time, in this case 100s, spent to perform a forward run by the subject along segment 11 , and then becomes constant, as the subject travels again through points that he / she has already visited.
[0093] Fig. 14a shows an exploration trajectory y” that is a circle 12 travelled along only once, whose centre coincides with the centre of region of interest 1. In other words, exploration trajectory y” follows the locus of points equidistant from a singular point, which is the centre of region of interest 1 , starting from an arbitrary point on the circle. The execution time T of exploration trajectory y” is 1000 units of time, for example 1000 s. The diagram in Fig. 14b shows the values of the subject’s distance from the fixed point and the score calculated over time according to the method of the invention and to the prior art, more precisely, by sampling k instants (ti ...tk) and applying the method of the invention / the prior art to calculate spatial exploration for each instant tk. It can be observed that, as the exploration proceeds, the score based on the distance from a fixed point, according to the prior art, remains unchanged, since the distance from the centre of the region of interest does not change, whereas the score according to the method of the invention increases progressively until the trajectory is completed (circle 12), even in this case in a manner that is more consistent with the actual exploration of region of interest 1 , and afterwards the score change become minimal, on the right-hand side of the diagram, since from that point onwards the trajectory approaches points that have already been visited.
[0094] Fig. 15a shows an exploration trajectory y”’ that starts at a point 15 and includes a path 13 comprising a first portion 16 running parallel to the entire boundary of region of interest 1 , and a second spiral portion 17 with a centre coinciding with the centre of region of interest 1 , whereby the subject first travels close to the boundaries of region of interest 1 and then visits central areas of region of interest 1 up to the centre, thus covering a large portion of the available surface. The execution time T of exploration trajectory y”’ is 1000 time units, for example 1000 s. The diagram in Fig. 15b shows the values of the area engaged by the subject and the score according to the method of the invention calculated over time, more precisely by sampling k instants (ti ...tk) and applying the method of the invention I the prior art to calculate spatial exploration for each instant tk. The example in Figs. 15a and 15b thus consists of a comparison between the score calculated according to the invention and the score based on the bounding box of exploration trajectory y”’. Since exploration trajectory y”’ initially moves entirely around the edge of the space, the values for the covered area saturate immediately and then remain constant over time, neglecting theamount of movement that occurs towards the centre of the space. In contrast, the score according to the method of the invention shows an exploration value that increases over time, as the trajectory unfolds towards the centre of the region of interest.
[0095] In Figs. 13b, 14b and 15b, the left vertical axis shows the score according to the method of the invention, while the right vertical axis shows, in arbitrary units, respectively, the distance travelled, the distance from a fixed point, in this case the centre of region of interest 1 , and the area engaged by the exploring subject, i.e. the three variables on which the aforementioned families of methods for calculating spatial exploration are based.
[0096] A test was also conducted on a sample of five male children aged between 1.7 and 9.5 years, who were asked to play in a 4.80 x 3-metre environment configured as a playground, where they could move freely and independently. Each child corresponds to a single exploration trajectory. Each exploration trajectory was captured by means of two distinct data acquiring systems: a) Nexus 3D Vicon, 100 Hz, 8 Vero video cameras, based on three retro-reflective markers positioned on the children’s foreheads; b) A GoPro Hero 4, mounted on the ceiling at the centre of the environment, 3 m above the floor. The camera recorded at a resolution of 1920x1080, with a Wide Field of View setting (20 fps, 120° field of view). The trajectories were extracted from the video stream by Segment Anything 2 model.
[0097] Table 1 shows the scores obtained using the method of the invention for both data acquiring systems.
[0098] Fig. 16 shows a non-parametric t-test for paired samples used to statistically compare the results, from which it can be concluded that there were no statistically significant differences between the scores obtained with the twoacquiring systems (t= -0.31857, p=0.77), demonstrating the invariance of the method according to the invention with respect to the trajectory acquiring system, with or without markers.
[0099] The above description of specific embodiments of the invention will be able to illustrate the invention from a conceptual point of view, such that others, utilising the prior art, can modify and / or adapt these specific embodiments in various applications without further research and without departing from the inventive concept and, therefore, it is understood that such adaptations and modifications shall be considered equivalent to the specific embodiments. The means and materials for carrying out the various functions described can be of various kinds without thus departing from the scope of the invention. It is understood that the expressions or terminology used are purely descriptive and, therefore, not limitative.
Claims
CLAIMS1. A method for quantifying a spatial exploration of a region of interest (1) carried out by a subject (3) through an exploration trajectory (y), said method comprising the steps of:prearranging (S10) an electronic computer (30) including a data input unit (31) and a calculation unit (32);defining (S20) a system of spatial coordinates (x,y) for said region of interest (1);delimiting (S30) said region of interest (1) in said spatial coordinates; detecting (S40) said exploration trajectory (y) by sampling (S45), at consecutive time instants (tk, k=1 ...M) of an execution time (T) of said exploration trajectory (y), positions (Gk=G(tk), k=1...M) engaged by said subject (3) in said region of interest (1);acquiring (S50), by said calculation unit (32), data of said exploration trajectory (y) in the form of position coordinates (x(tk), y(tk), k=1 ...M) of said positions (Gk=G(tk), k=1 ...M);acquiring (S60), by said calculation unit (32), node coordinates (xi,yi, i=1...N) of a number (N) of nodes of a predetermined mesh (2) of nodes (Pi, i=1 ...N) lying within said region of interest (1); determining (S70), by said calculation unit (32), a distance (n) of each node (Pi) of said nodes of said mesh (2) from a position (Gk ), among said positions (Gk, k=1 ...M), closest to said each node (Pi), obtaining a plurality of raw node-trajectory distance values (n, i=1 ...N); calculating (S80), by said calculation unit (32), a spatial exploration indicator starting from said plurality of raw node-trajectory distance values (n, i=1 ...N).
2. The method according to claim 1 , wherein said step of calculating (S80) a spatial exploration indicator by said calculation unit (32) comprises a step of calculating a spatial exploration score in the form of a raw spatial exploration score (S) of said exploration trajectory (y) as an average of said raw node-to-trajectory distance values (n, i=1 ...N).
3. The method according to claim 1 , wherein said step of calculating (S80) a spatial exploration indicator by said calculation unit (32) comprises, aftersaid step (S70) of determining a distance (n, i=1 ...N) of each node (Pi) from said exploration trajectory (y), a step ofgenerating (S90) a heatmap in said system of spatial coordinates (x,y), by said calculation unit (32) and starting from said plurality of raw node-trajectory distance values (n, i=1 ...N), said heatmap comprising, at nodes (Pi) of said mesh (2), respective coloured areas each having a colorimetric feature or a symbol repetition density monotonically depending on said respective distance values (n, i\\=1 ...N).
4. The method according to claim 1 , whereinsaid step of delimiting (S30) said region of interest (1 ) comprises a step of:acquiring (S35), by said calculation unit (32), intervals (Ax, Ay) of said spatial coordinates (x,y) representing said region of interest (1), and said region of interest (1) has a polygonal shape with a number (H) of vertices, andsteps are provided of:inputting (S33) vertex coordinates (xh,yh, h=1 ...H) of said number (H) of vertices (Vh, h=1 ...H) of said polygonal shape into said data input unit (31);transferring (S34) said vertex coordinates (xh,yh, h=1 ...H) from said data input unit (31) to said calculation unit (32), andsaid step (S35) of acquiring said spatial coordinate intervals (Ax, Ay) representing said region of interest (1) by said calculation unit (32) includes a step of calculating (S36) said spatial coordinate intervals (Ax, Ay) representing said region of interest (1), by said calculation unit (32), and starting from said vertex coordinates (xh,yh, h=1 ...H).
5. The method according to claim 4, wherein said number (H) of vertices is 4 and said polygonal shape is a rectangle.
6. The method according to claim 1 , whereinsaid step of delimiting (S30) said region of interest (1) comprises a step of acquiring (S35), by said calculation unit (32), intervals (Ax, Ay) of said spatial coordinates (x,y) representing said region of interest (1), and steps are provided of:inputting (S31) said intervals (Ax, Ay) of said spatial coordinates (x,y) representing said region of interest (1) into said data input unit (31); transferring (S32) said intervals (Ax, Ay) of said spatial coordinates (x,y) representing said region of interest (1) from said data input unit (31) to said calculation unit (32), so as to cause said step (S35) of acquiring said intervals (Ax, Ay) to be performed by said calculation unit (32).
7. The method according to claim 1 , wherein said step (S60) of acquiring said node coordinates (xi,yi, i=1 ...N) includes a step of calculating (S65) said node coordinates (xi,yi, i=1 ...N) by said calculation unit (32).
8. The method according to claim 1 , wherein said mesh (2) is a matrix of said nodes (Pi, i=1 ...N) spaced apart from each other by a predetermined pitch (P)- 9. The method according to claims 7 and 8, whereinbefore said step (S60) of acquiring said node coordinates (xi,yi, i=1 ...N), the following steps are performed:said steps of detecting (S40) said exploration trajectory (y) and of acquiring (S50) said data of said exploration trajectory (y); steps of:calculating (S57) partial lengths (zk, k=1...M) of consecutive portions (yk, k=1 ...M) of said exploration trajectory (y) between consecutive positions (Gk,Gk+i k=1 ...M) sampled along said exploration trajectory (y) during said execution time (T) of said exploration trajectory (y), thus obtaining a partial length distribution;extracting (S58) a statistic of said partial length distribution, selected between the maximum value (zmax) and the upper fence of said partial length distribution; andin said step (S65) of calculating said node coordinates (xi,yi, i=1 ...N) by said calculation unit (32), said pitch (p) is selected (S59) not exceeding said statistic of said partial length distribution.
10. The method according to claim 9, wherein said pitch (p) is selected substantially equal to said statistic of said partial length distribution (zk, k=1...M).
11. The method according to claim 1 , wherein steps are provided of:inputting (S61) said node coordinates (xi,yi, i=1 ...N) into said data input unit (31);transferring (S62) said node coordinates (xi,yi, i=1 ... N) from said data input unit (31) to said calculation unit (32), so as to cause said step (S60) of acquiring said number (N) of said node coordinates (xi,yi, i=1 ...N) to be performed by said calculation unit (32).
12. The method according to claim 1 , whereinsaid exploration trajectory (y) is physically travelled by said subject (3) in said region of interest (1), during said execution time (T), andsaid subject (3) is selected from the group comprised of:a human being;an animal;a microorganism;a vehicle or moving equipment selected from the group comprised of: - a motor vehicle;an airplane;a drone;a boat;a robot, in particular a robotic vacuum cleaner or floor cleaner; - a robotic lawnmower.
13. The method according to claim 12, wherein said step (S40) of detecting said exploration trajectory (y) is performed by means of a device selected from the group comprised of:a device comprising a plurality of video cameras (6q, q=1 ...Q), each having a field of view (4q) including a respective portion of said region of interest (1);a device comprising a single video camera (6) having a field of view (4) including said region of interest (1);a microorganism tracking device incorporated in a microscope;a device comprising a tracer (8) of a predetermined sensor (7), said tracer (8) configured to be worn / installed on said human or animal I on said moving equipment, respectively, and having a reception field (5) including said region of interest (1).
14. The method according to claim 1 , whereinsaid region of interest (1 ) is a portion of a scenario observed by said subject (3), andsaid exploration trajectory (y) is drawn on said region of interest (1) by an optical axis oriented by said subject (3) through his / her own eye movements, andsaid step (S40) of detecting said exploration trajectory (y) is performed by an eye movement tracking device.
15. The method according to claim 1 , whereinsaid region of interest (1) is a portion of a screen of a computer or similar device controlled by said subject (3) through a pointer means, and said step (S40) of detecting said exploration trajectory (y) is performed by a cursor tracking device for tracking a position on said screen of a cursor associated with said pointer means.
16. The method according to claim 2, wherein a step is further provided of:normalising (S85) of said raw spatial exploration score (S) so as to obtain a normalised spatial exploration score (S) set between a minimum score value and a maximum score value.
17. The method according to claim 16, wherein said minimum score value and said maximum score value of said normalised spatial exploration score (S) are 0 and 1 , respectively.
18. The method according to claim 16, wherein said minimum score value and said maximum score value correspond to minimum and maximum exploration conditions, respectively, of said region of interest (1).
19. The method according to claims 4 and 16, wherein said spatial exploration score is calculated as a normalised spatial exploration score (S) by a formula selected between:andwhereS is said normalised space exploration score;S is said raw spatial exploration score;Smaxis a maximum raw spatial exploration score, calculated by the steps of:for each vertex of said vertices (Vh, h=1 ...H) of said polygonal shape of said region of interest (1):calculating a distance of said each vertex (Vh) from each node (Pi, i=1 ...N) of said number (N) of nodes (Pi, i=1 ...N) of said mesh (2), thus obtaining a plurality of vertex-node distances;calculating an average vertex-node distance between each of said vertices (Vh) and said nodes (Pi, i=1 ...N); thus obtaining a plurality of average vertex-node distances; defining said maximum raw spatial exploration scoreas the maximum value among said average vertex-node distances.
20. A system (10) for quantifying a spatial exploration of a region of interest (1 ) carried out through an exploration trajectory (y) by a subject (3), said system (10) comprising:a detector device (20) configured to sample, at consecutive time instants (tk, k=1 ...M) of an execution time (T) of said exploration trajectory (y), positions (Gk=G(tk), k=1 ...M) engaged by said subject (3) in said region of interest (1) while traveling along said exploration trajectory (y);an electronic computer (30) including a data input unit (31 ) and a calculation unit (32),wherein said calculation unit (32) is configured to:delimit said region of interest (1 ) by means of a system of spatial coordinates (x,y);acquire node coordinates (xi,yi, i=1 ...N) of a number (N) of nodes of a predetermined mesh (2) of nodes (Pi, i=1 ...N) lying within said region of interest (1);acquire data of said exploration trajectory (y) in the form of position coordinates (x(tk), y(tk), k=1 ...M) of said positions (Gk=G(tk), k=1 ...M);measure a distance (n) of each node (Pi) of said mesh (2) from a position (Gk ), among said positions (Gk, k=1 ...M), closest to said each node (Pi), obtaining a plurality of raw node-trajectory distance values (n, i=1...N);calculate a spatial exploration indicator of said exploration trajectory (y) starting from said plurality of raw node-trajectory distance values (n, i=1 ...N).
21. The system according to claim 20, wherein said calculation (S80), for which said calculation unit (32) is configured, of a spatial exploration indicator comprises calculating a spatial exploration score in the form of a raw spatial exploration score (S) as an average value of said raw nodetrajectory distance values (n, i=1 ...N).
22. The system according to claim 20, wherein said calculation (S80), for which said calculation unit (32) is configured, of a spatial exploration indicator comprises, starting from said plurality of raw node-trajectory distance values (n, i=1 ...N), comprises a generation of a heatmap in said system of spatial coordinates (x,y), said heatmap comprising, at nodes (Pi) of said grid (2), respective colored areas each having a colorimetric feature or a symbol repetition density monotonously depending on respective said raw node-trajectory distance values (n, i=1 ...N).
23. The system (10) according to claim 20, wherein said region of interest (1) has a polygonal shape with a number (H) of sides, andsaid data input unit (31) is configured to receive and transfer to said calculation unit (32) vertex coordinates (xh,yh, h=1 ...H) of a plurality of vertices (Vh, h=1 ...H) of said polygonal shape;said calculation unit (32) is configured to calculate said spatial coordinate intervals (Ax, Ay) representative of said region of interest (1) starting from said vertex coordinates (xh,yh, h=1 ...H).
24. The system (10) according to claim 20, wherein said calculation unit (32) is configured to calculate said node coordinates (xi,yi, i=1 ...N).
25. The system (10) according to claim 20, wherein said calculation unit (32) is configured to calculate said node coordinates (xi,yi, i=1 ...N) of nodes (Pi, i=1 ...N) of a matrix having a predetermined pitch (p) between one another.
26. The system (10) according to claims 24 and 25, wherein said calculation unit (32) is configured tosample consecutive positions (Gk,Gk+i k=1...M) on said exploration trajectory (y) during said execution time (T) of said exploration trajectory (y);calculate partial lengths (zk, k=1...M) of consecutive portions (yk, k=1...M) between consecutive positions (Gk,Gk+i k=1...M) thus obtaining a partial length distribution;extract a statistic of said partial length distribution selected between the maximum value (zmax) and the upper fence of said partial length distribution;calculate said node coordinates (xi,yi, i=1 ... N), selecting said pitch (p) as a value not exceeding said statistic of said partial length distribution, in particular as a value substantially equal to said statistic of said partial length distribution.
27. The system (10) according to claim 20, wherein said detector device (20) is selected from the group comprised of:a device comprising a plurality of video cameras (6q, q=1 ...Q) each having a field of view (4q) including a respective part of said region of interest (1);a device comprising a single camera (6) having a field of view (4q) including said region of interest (1);a microorganism tracking device incorporated in a microscope;a device comprising a tracer of a predetermined sensor (S) configured to be worn / installed on a human or animal being / on a mobile device, respectively, and having a reception field (5) including said region of interest (1);- a human or animal being’s eye movement tracking device;a cursor tracking device for tracking a position of a cursor associated with a pointer means on said screen.
28. The system (10) according to claim 20, wherein said calculation unit (32) is configured to normalize said raw spatial exploration score (S) so as to obtain a normalized spatial exploration score (S) between a minimum value and a maximum value, in particular between 0 and 1 ,in particular, said minimum and maximum values correspond to minimum and maximum exploration conditions, respectively, of said region of interest (1) by said subject (3).
29. The system (10) according to claim 28, wherein said calculation unit (32) is configured to calculate said normalized spatial exploration score (S) using a formula selected betweenandwheren, i=1... N are said raw node-trajectory distance values;(S) is said raw spatial exploration score;Smaxis the maximum raw spatial exploration score,S is said normalized spatial exploration score (S).
