METHOD FOR CONFIGURING AN OPTOELECTRONIC SENSOR

Description

[0001] The invention relates to a method for configuring an optoelectronic sensor according to the preamble of claim 1.

[0002] For the light receiver of an optoelectronic sensor, a region of interest (ROI) is defined in many applications. This involves selecting a sub-area within the typically matrix-shaped arrangement of light-receiving elements. In image processing, this is a common concept for reducing the amount of data. For a tactile sensor that emits and receives a light beam, the region of interest can be defined so that the light spot generated by the beam is captured. Light-receiving elements not touched by the light spot would only contribute ambient light and thus ultimately an interference signal.

[0003] Individual light-receiving elements or cells of a light receiver can be defective or exhibit undesirable properties, for example, being too light-sensitive ("screamer") or too insensitive ("dark pixel"). If defective cells fall within the area of ​​interest, they distort the measurement result. Conventional light transmitter alignment only ensures that the light spot falls on the light receiver. Whether defective cells are encountered is not checked and, given the statistical distribution of defective cell positions, is also unpredictable. If too many defective sensor cells are present in the area of ​​the light spot or the area of ​​interest, the specified sensor properties are not achieved, and the device may have to be discarded.

[0004] Of course, it's possible to switch to a manufacturing technology where defective cells occur less frequently. However, this hardly avoids costs, as the increased yield comes at the expense of significantly more expensive light receivers. Alternatively, defective cells can simply be permitted in the relevant area. In that case, however, a greater variation in sensor performance must be expected. Accordingly, all sensors should be specified with a lower performance level as a precaution, in order to maintain a buffer for variance. This ensures that the technical limits are not fully exploited.

[0005] To detect even low light intensities, some optoelectronic sensors employ avalanche photodiodes (APDs). The incident light triggers a controlled avalanche breakdown (avalanche effect). This multiplies the charge carriers generated by the incident photons, resulting in a photocurrent that is proportional to the light intensity but significantly larger than that of a simple PIN diode.

[0006] Even greater sensitivity is achieved with avalanche photodiodes operated in so-called Geiger mode (SPAD, Single Photon Avalanche Diode). In this mode, the avalanche photodiode is biased above its breakdown voltage, so that a single charge carrier released by a single photon can trigger an uncontrolled avalanche. Due to the high field strength, this avalanche then recruits all available charge carriers. Afterward, the avalanche comes to a halt (passive quenching) and is no longer detectable for a certain dead time. Alternatively, it is also known to detect and extinguish the avalanche externally (active quenching).

[0007] A SPAD thus counts individual events, much like a Geiger counter. SPADs are not only highly sensitive but also relatively inexpensive and efficient to integrate into silicon semiconductors. Furthermore, they can be easily integrated onto a circuit board. A key feature is that even a minimal interference event, such as an extraneous light photon or dark noise, generates the same maximum received signal as a useful light signal.

[0008] Distance sensors or distance probes based on the time-of-flight principle measure the travel time of a light signal, which corresponds to the distance via the speed of light. A distinction is made between pulse-based and phase-based measurement. In a pulse-time-of-flight method, a short light pulse is emitted, and the time until the reception of a remission or reflection of the light pulse is measured. Alternatively, in a phase-based method, the transmitted light is amplitude-modulated, and a phase shift between the transmitted and received light is determined, with the phase shift also serving as a measure of the light travel time.

[0009] A SPAD-based distance sensor preferably operates using pulse-based technology to ensure robust measurements even when encountering edges or during remission jumps. This is also known as direct time-of-flight (dToF) measurement. To obtain reliable measurement results despite interference, events can be collected across multiple SPADs or multiple transmission pulses and evaluated together by searching for a maximum in a histogram.

[0010] From EP 3 428 683 B1, an optoelectronic sensor is known that measures distances using SPADs and a dToF method. A switch matrix is ​​provided to selectively choose specific SPADs for evaluation and connect them one-to-one with TDCs (Time-to-Digital Converters). A single, combined histogram is then collected from the selected SPADs. The document does not address the issue of defects in the selected cells or SPADs.

[0011] In US 8 040 406 B2, defective pixels are compensated for using image information from neighboring pixels. This compensation does not change the fact that the actual information at the pixel's position is lost. While US 8 040 406 B2 does not recognize any areas of interest, US 2022 / 0360673 A1 proposes correction within a region of interest (ROI). However, this relies on data formats with metadata that a simple sensor, for example, for distance measurement, does not use. Furthermore, limiting the approach to a single ROI does not alter the fact that such compensation is an interpolation that merely masks the defect. In JP 2013 239772 A, movements within a single ROI are reconstructed from multiple images despite defective pixels. This requires significant computational effort and relies on multiple exposures. Each individual detection remains affected by the defective pixels.

[0012] It is therefore an object of the invention to further improve the detection of a sensor when using an area of ​​interest.

[0013] This problem is solved by a method for configuring an optoelectronic sensor according to claim 1. The light receiver comprises a plurality of light-receiving elements or cells, arranged, for example, in a matrix. A control and evaluation unit is at least indirectly connected to the light receiver and can accordingly read signals from the light-receiving elements or information derived therefrom. A selection of a portion of the light-receiving elements is made to form a region of interest (ROI). Only the information from the light-receiving elements of the region of interest is further evaluated in the control and evaluation unit; preferably, only this information is read from the light receiver. Information from the light-receiving elements outside the region of interest is therefore ignored, at the latest during the evaluation.This can relate to a specific function of the sensor, for which an area of ​​interest is selected, or preferably, the light receiver is limited to the area of ​​interest in such a way that the remaining light-receiving elements can be considered inactive after the configuration is complete, at least from a functional point of view. The area of ​​interest can be continuous, particularly in a simple geometric form such as a rectangle, a circle, or a line, but this need not be the case, as in the case of several separate geometric shapes, such as two rectangles or circles, or a grid over at least a sub-area of ​​the light receiver.

[0014] The invention is based on the fundamental idea of ​​selecting the area of ​​interest in such a way that defective light-receiving elements are excluded, at least to a tolerated maximum number. To this end, the defective light-receiving elements are first identified. A light-receiving element is considered defective, in particular, if it delivers a signal regardless of the incident light or reacts too sensitively or too insensitively to incident light ("stuck-at", "screamer", "dark pixel"). The area of ​​interest is then positioned or shifted so that at most a maximum number of defective light-receiving elements are selected for the area of ​​interest, thus avoiding the selection of defective light-receiving elements for the area of ​​interest. In simpler terms, the area of ​​interest is placed within an intact area of ​​the light receiver that is not affected by defective light-receiving elements.The maximum permissible number of defective light-receiving elements, which directly corresponds to a maximum proportion given the size of the area of ​​interest, is a selectable parameter.

[0015] The execution of the specified process steps is automated, preferably within the sensor and / or a connected configuration or diagnostic device. This may involve an instruction to modify the sensor, preferably during its manufacture, for example, to relocate the light receiver or place it at a location specified in the instruction. This can then be followed by another fully automated step that verifies the execution of the instruction and, in particular, checks whether it is now possible to select an area of ​​interest with the maximum number of defective light-receiving elements.

[0016] The invention has the advantage that, by appropriately selecting a region of interest, a sensor can remain fully functional despite defective light-receiving elements. This increases the proportion of sensors that meet the requirements of a detection task, and in particular the yield during manufacturing. It is no longer necessary to artificially limit the specified performance to ensure consistent behavior across all sensors. The number of defective light-receiving elements in the region of interest no longer depends on their statistical distribution; conversely, the region of interest is positioned in such a way that such statistical errors are compensated for.Statistical outliers with an unusually high number and / or unusually unfavorable distribution of defective light-receiving elements remain possible, but only with a significantly more favorable quantile of devices to be sorted out.

[0017] The area of ​​interest is preferably selected such that it does not include any defective light-receiving elements. A particularly stringent condition is formulated here, allowing absolutely no tolerance for defective light-receiving elements in the area of ​​interest. This ensures consistent, full functionality across all configured sensors. Such a strict yield requirement can be imposed because the presence of a sufficiently small number of defective light-receiving elements is largely no longer a matter of chance. On the contrary, the light receiver may well exhibit defects, which are avoided according to the invention.

[0018] The selection of the area of ​​interest is preferably carried out during the sensor's manufacturing process. Generally, the determination of the area of ​​interest can be done during configuration, for example, on-site commissioning. However, according to this embodiment, this is already part of the manufacturing process; the method for configuring an optoelectronic sensor thus becomes a method for manufacturing an optoelectronic sensor, in which the selection of the area of ​​interest is a manufacturing step. If no area of ​​interest can be selected with at most the maximum number of defective light-receiving elements, the sensor is discarded or, if necessary, marketed as a sensor with a lower specified performance.

[0019] Defective light-receiving elements are preferably detected during the sensor's acquisition of a reference object. Thanks to the reference object, whose remission properties are defined, there is an expectation for the respective signals of the light-receiving elements. These can be predefined as data for the reference object used, or an expectation can be created for each light-receiving element by comparison with its neighboring light-receiving elements. A defined illumination or darkness within the detection area can also preferably serve as a reference object, creating a reference scenario without directly detecting a specific reference object.

[0020] The sensor preferably comprises a light emitter whose emitted light creates a light spot on the light receiver, wherein the area of ​​interest is selected such that the light spot falls on light-receiving elements of the area of ​​interest. This embodiment relates to the large class of tactile optoelectronic sensors that evaluate a reflection of their own emitted light. In this case, the area of ​​interest and the light spot are linked and can, in a certain sense, be considered synonymous. Preferably, the area of ​​interest is centered on the light spot. Various degrees of overlap between the area of ​​interest and the light spot are possible, in particular a complete overlap in which the area of ​​interest comprises precisely the light-receiving elements illuminated by the light spot.However, it is also conceivable that at least some of the light-receiving elements illuminated by the light spot do not belong to the area of ​​interest, or conversely, that the area of ​​interest is somewhat too large and contains at least some light-receiving elements that the light spot does not hit.

[0021] Preferably, the position of the light spot on the light receiver is shifted by moving the light source, a transmitting optic of the light source, the light receiver, and / or a receiving optic of the light receiver. These are mechanical measures to achieve a new light spot position if there are too many defective light-receiving elements in the vicinity of the previous light spot position. The corresponding components are moved or tilted, particularly during manufacturing before fixation, such as an adhesive process. The area of ​​interest is functionally bound to the light spot and therefore cannot be meaningfully changed independently of the light spot position. Without changing the light spot position, the area of ​​interest could avoid defective light-receiving elements, but would no longer detect the light spot, or at least not completely.

[0022] From the identified defective light-receiving elements, a defect map of the light receiver is preferably generated, highlighting the positions of the defective elements. The defect map is, for example, a matrix corresponding to the arrangement of the light receiver's elements. The entries on the defect map, using a user-defined code, indicate whether the light-receiving element at that position is defective or not, optionally with more detailed information about the type of defect. This allows the defect map to provide quick access to all relevant information about the defective light-receiving elements.

[0023] The defect map is preferably folded with a region of interest (ROI) map corresponding to the shape of the region of interest to form an overlap map. The ROI map is smaller than the defect map and specifies the desired arrangement of light-receiving elements of the region of interest. For example, it is a matrix with the dimensions of the region of interest. Folding the ROI map with the defect map creates an overlap map. From the overlap map, it can be determined for each position how many defective light-receiving elements would be located in a region of interest arranged at that position.

[0024] The ROI map preferably has a buffer zone around the area of ​​interest. The ROI map is therefore slightly larger than the actual area of ​​interest in at least one direction, and preferably in several or all directions. This ensures greater robustness against manufacturing tolerances, for example, from an adhesive bonding process that fixes the relative positions within the sensor. If the displacements remain smaller than the extent of the buffer zone, defective light-receiving elements initially only reach the buffer zone due to the tolerances, but not yet the area of ​​interest. The buffer zone can only fully accommodate such tolerances if it is sufficiently large. The buffer zone is preferably incorporated into the folding described in the preceding paragraph in such a way that it remains clear whether a defective light-receiving element overlaps with the area of ​​interest or the buffer zone.This allows for better optimization later on, because a defective light receiving element in the buffer zone is usually more acceptable than one in the area of ​​interest itself.

[0025] Positions on the overlap map are preferably evaluated using a defect measure, where the defect measure takes on a value depending on the overlap of the ROI map with defective light-receiving elements. Positions without overlap with defective light-receiving elements are particularly favorable and therefore have a low defect measure. Each defective light-receiving element that would be located in an area of ​​interest at the considered position increases the defect measure. The contribution of a defective light-receiving element in a buffer zone to the defect measure preferably remains lower.

[0026] The area of ​​interest is preferably selected such that the defect measure for the positions of the light-receiving elements within that area does not exceed a maximum defect measure. Ideally, the defect measure remains zero, meaning the area of ​​interest contains no defective light-receiving elements at all. Whether a lower maximum defect measure is acceptable depends on the sensor and its application. It is conceivable that the light receiver, due to too many defective light-receiving elements and / or their unfavorable distribution, does not offer a suitable position. In this case, configuration is not possible, and the sensor must be rejected or operated at a reduced performance level, which, unlike in the prior art, is then known and can be specified.Conversely, there may be several suitable positions, among which a decision is then made randomly or according to further criteria, such as minimizing the required sensor modification. The numerical conventions for the defect measure are selectable. In particular, it is assumed without loss of generality that a favorable defect measure remains small; an inverted convention with a high favorable defect measure ultimately leads to the same results.

[0027] The defect size is preferably reduced at positions within a predefined area of ​​the light receiver. This predefined area represents a preference for desired positions on the light receiver, externally determined. The defect size is preferably only slightly reduced in this area to prioritize defects higher than this predefined area. However, the reverse scenario, with a significant reduction in the defect size, is also conceivable. In this case, the predefined area should be adhered to under all circumstances, even at the risk of defective light-receiving elements being located within the area of ​​interest, or even if no suitable area of ​​interest is found outside the predefined area, where an area of ​​interest with no or fewer defective light-receiving elements would have been possible.

[0028] The defect measure preferably includes a component that depends on the distance of the position evaluated with the defect measure from an initial position of the area of ​​interest, in particular an initial light spot position. It is assumed here that there is already an initial preferred position of the area of ​​interest, for example, where a light emitter of the sensor generates a light spot. A distance-dependent defect measure then favors smaller changes. The distance dependency can be non-linear and, in particular, may disproportionately penalize or completely prevent larger changes.

[0029] The light-receiving elements are preferably configured as avalanche photodiode elements in Geiger mode. Such light-receiving elements, also known as SPADs, were briefly described in the introduction and make the light receiver particularly sensitive.

[0030] The sensor preferably comprises a plurality of time-of-flight measurement units that can be variably connected to light receiving elements, with the selection of an area of ​​interest being achieved by the connection between the light receiving elements and the time-of-flight measurement units. Here, individual light receiving elements or groups of them are each connected to a time-of-flight measurement unit that measures a single light transit time. The time-of-flight measurement units can, for example, include a time-to-digital converter (TDC). The individual light transit times are collected, for example, in a memory, and the control and evaluation unit generates a distance value from the collected individual light transit times. There are fewer time-of-flight measurement units than light receiving elements.This means that only a portion, or even just a small portion, of the light-receiving elements are connected to a time-of-flight measuring unit, and the selected light-receiving elements constitute the area of ​​interest. The selection of light-receiving elements, or their assignment to time-of-flight measuring units, is particularly preferably carried out using a switch matrix. Further details regarding the design of time-of-flight measuring units, the measurement of the time of light, and the variable selection and assignment of light-receiving elements to time-of-flight measuring units can be found in the aforementioned EP 3 428 683 B1, whereby the invention preferably uses one-to-one assignments between light-receiving elements and time-of-flight measuring units.

[0031] The invention is further explained below with regard to additional features and advantages by way of example embodiments and with reference to the accompanying drawing. The illustrations in the drawing show: Fig. 1 a schematic representation of an optoelectronic sensor; Fig. 2 an illustration of a defect map of a light receiver showing the positions of defective light-receiving elements; Fig. 3 an illustration of a region of interest (ROI) map showing a planned arrangement of light-receiving elements of a region of interest, including a buffer zone; Fig. 4 an illustration of folding a defect map with an ROI map to create an overlap map; Fig. 5 a similar illustration Figure 4Figure 1 shows only the folding of a defect map with the buffer zone of an ROI map; Figure 6 is an illustration of a combined overlap map; Figure 7 is an illustration of an overlap map evaluated with a defect measure and a new light spot position derived therefrom; Figure 8 is an illustration of an area of ​​interest corresponding to the new light spot position, wherein the area of ​​interest does not contain any defective light-receiving elements; and Figure 9 is a comparative representation of the increased yield of sensors due to the shift of the area of ​​interest while avoiding defective light-receiving elements. Figure 1 Figure 1 shows a schematic representation of an optoelectronic sensor 10. A light emitter 12, for example an LED or a laser light source, emits a light signal 16 into a detection area 18 via a transmitting optic 14. If the light signal 16 encounters an object 20 in the detection area 18, a remitted or reflected light signal 22 returns via a receiving optic 24 to a light receiver 26. This light receiver 26 has a plurality of light receiving elements 28, which are preferably configured as Geiger-mode avalanche photodiode elements or SPADs and which can be considered sensor cells or pixels. SPADs provide a virtually digital signal and thus react extremely quickly to incoming light.

[0032] To detect even weakly reflective objects 20 and achieve a long range, the optical front end with transmitting optics 14 and receiving optics 24 is designed so that as much light as possible reaches the light receiver 26. However, the light spot generated on the light receiver 26 by the incident reflected light signal 22 typically only falls on a portion of the light-receiving elements 28 around the light spot position 30. In this sense, there are too many light-receiving elements 28. In a configuration of the sensor 10, preferably during manufacturing, the appropriate light-receiving elements 28 are selected, i.e., preferably those onto which the light spot falls. The selected light-receiving elements 28 form a region of interest 32 (ROI). The signals from the remaining light-receiving elements 28 outside the region of interest 32 are not read out or at least not further evaluated.Unselected light receiving elements 28, whose signals would not be used anyway, can be switched off, for example, by lowering the bias voltage below the breakdown voltage.

[0033] A control and evaluation unit 34 is responsible for controlling the various components of the sensor 10 and evaluating the signals from the selected light receiving elements 28 within the area of ​​interest 32. The control and evaluation unit 34 comprises at least one processing unit. Examples of such processing units are digital computing devices such as a microprocessor or CPU (Central Processing Unit), an FPGA (Field Programmable Gate Array), a DSP (Digital Signal Processor), an ASIC (Application-Specific Integrated Circuit), a K-processor, an NPU (Neural Processing Unit), a GPU (Graphics Processing Unit), or the like.

[0034] The evaluation can only yield a binary object detection signal indicating whether the light signal 16 has encountered an object 20 or not. Preferably, the sensor 10 is capable of measuring the distance to a detected object 20 using a time-of-flight method. The emitted light signal 16 then preferably has a light pulse, so that the sensor 10 determines distances using the pulse method or a direct time-of-flight method (dToF). From the time of flight between the emission of a light signal 16 and the reception of the corresponding reflected light signal 22, the distance to the object 20 can be deduced using the speed of light.

[0035] In a preferred statistical evaluation, individual light transit times are determined for each of the selected light receiving elements 28 of the area of ​​interest 32 and accumulated, for example, in a histogram. A plurality of light transit time measurement units, in particular TDCs (Time-to-Digital Converters), can be provided for determining individual light transit times. In this context, an advantageous specific form of selecting the area of ​​interest 32 results from assigning selected light receiving elements 28 to a limited number of light transit time measurement units. Reference is also made to EP 3 428 683 B1 mentioned in the introduction. The control and evaluation unit 34 is preferably configured to further process histograms from individual light transit times and, for example, to locate a peak caused by the reflected light signal 22 that corresponds to the time of reception.It is conceivable to integrate at least some or parts of the light time-of-flight measuring units and / or functional blocks of the control and evaluation unit 34 with the light receiving elements 28 on a common chip (e.g. ASIC, Application-Specific Integrated Circuit).

[0036] The arrangement of sensor 10 in Figure 1 This is purely an example. Other known optical solutions can be used alternatively, such as autocollimation with a beam splitter and a common optic, or placing the light source 12 in front of the light receiver 26. More complex sensors such as light grids or laser scanners are also conceivable.

[0037] In a real light receiver 26, it regularly occurs that at least some of the light-receiving elements 28 are defective. According to the invention, the area of ​​interest 32 is defined such that it contains no defective light-receiving elements 28 or only a limited number of defective light-receiving elements 28. This definition of the area of ​​interest 32 preferably takes place during the manufacturing process. Before this is subsequently carried out with reference to the Figures 2 to 8 A brief overview will be given, as the procedure will be described in detail using an exemplary approach.

[0038] As a starting point or input parameter for optimizing the area of ​​interest 32 with regard to defective light-receiving elements 28, the defective light-receiving elements 28 are identified, for example, during the detection of a known reference object that provides expectations for the individual light-receiving elements 28, with which they can be compared to locate defects. A defined illumination or darkness in the detection area 18 preferably also serves as the reference object. Furthermore, a pattern of the desired area of ​​interest 32 is specified, either through parameterization, configuration software, or an algorithmic approach, or by measuring light spots at several intervals.

[0039] It is conceivable to define the area of ​​interest 32 only once it is in a suitable position. However, in the following, without loss of generality, it is assumed that an initial area of ​​interest 32 already exists, which is defined without regard to defective light-receiving elements 28. This can be easily achieved by defining any initial position. In particular, there is no freedom of choice regarding the position of the area of ​​interest 32 if it is intended to detect a light spot generated by the reflected light signal 22, which occupies a light spot position 30 determined by the structural conditions. The aforementioned input parameters then also include the initial position of the as-yet-unoptimized area of ​​interest 32, or the light spot position 30.

[0040] Optional parameters can include a maximum permissible displacement of the area of ​​interest 32 from its original position, a specific sub-area of ​​the light receiver 26 to be utilized and / or a number of permitted defective light-receiving elements 28.

[0041] The inventive method then determines a new position of the area of ​​interest 32, or a displacement required to achieve it, such that at most the permissible number of defective light-receiving elements 28 lies in the new area of ​​interest 32. A tolerance or buffer zone can be taken into account in case the new position is not reached ideally in practice. If the area of ​​interest 32 is to contain the light spot position 30, this must be changed according to the new position. For this purpose, the light emitter 12, the transmitting optics 14, the receiving optics 24, and / or the light receiver 26 can be mechanically moved, i.e., shifted and / or tilted, for example, using suitable grippers during the adjustment phase of manufacturing, before the components are fixed, for example, by an adhesive process.The area of ​​interest 32 can usually be changed very easily and without mechanical intervention by altering a variable circuit, as in the case of the EP 3 428 683 B1, by appropriate other control, or simply by changing the evaluation, and adapted to the new light spot position 30.

[0042] Figure 2Figure 1 shows an illustration of a defect map of the light receiver 26. The defect map is a matrix corresponding to the matrix of the light receiver 26 with highlighted positions of defective light-receiving elements 28. In more detail, the upper left shows a submap with light-receiving elements 28a with excessive sensitivity ("Screamer"), the upper right shows a submap with light-receiving elements 28b with insufficient sensitivity or no signal ("Dark Pixel"), the lower left shows a superposition of the two submaps, and the lower right shows the superposition with further information, namely the previous light spot position 30 and a predefined target area 36 for the area of ​​interest 32.

[0043] The defective light-receiving elements 28a-b are preferably identified by placing a reference object in the detection area 18 and comparing the signals of the light-receiving elements 28 with an expected value that is predefined or derived, for example, from the surrounding area. The reference object can also be defined by illuminating the light receiver 26, in particular by illuminating the detection area 18 or by complete darkness. The previous light spot position 30 can be considered the original location of an area of ​​interest 32, for example, marking its center. The target area 36 defines a desired position for the area of ​​interest 32, which corresponds to the aforementioned optional parameter of a specific sub-area of ​​the light receiver 26 to be utilized.

[0044] Figure 3Figure 1 shows an illustration of a ROI map with a proposed arrangement of positions 38 for the light-receiving elements 28 of the area of ​​interest 32. The ROI map is thus a geometric shape specification for the area of ​​interest 32, whose optimized position still needs to be determined. The depicted rectangular shape is purely exemplary; the ROI map can specify any shape, with two separate rectangles shown to illustrate the possible variations.

[0045] Preferably, the ROI map includes a buffer zone with in Figure 3The further positions 40 shown in lighter colors are for light-receiving elements 28. The buffer zone serves to compensate for tolerances during the final displacement of the light spot position 30 or the area of ​​interest 32 into the determined optimal position, for example, tolerances from an adhesive process or other fixing. The buffer zone optionally ensures that an area of ​​interest 32 slightly displaced from the optimal position has no or only a maximum number of defective light-receiving elements 28, as long as the displacement remains within the buffer zone. A wider buffer zone accommodates larger tolerances but makes it more difficult to find a suitable position without defective light-receiving elements 28 for the area of ​​interest 32 enlarged by the buffer zone, so a trade-off should be made here.It is possible to give less weight to defective light receiving elements 28 in the buffer zone when evaluating for finding an optimal position than in the actual area of ​​interest 32.

[0046] Figure 4 shows an illustration of a folding of the defect map according to Figure 2 with the ROI map according to Figure 3 The result is an overlap map. Positions 42 where an overlap occurs are highlighted, and preferably, the degree of overlap is shown. In other words, defective light-receiving elements 28 are located within an area of ​​interest 32 shifted to these positions 42. The buffer zone of the Figure 3 remains in Figure 4 disregarded.

[0047] Figure 5 shows an illustration similar Figure 4 , shown here is the folding of a defect card according to Figure 2 using only the buffer zone of the ROI map according to Figure 3. Therefore, defective light receiving elements 28 are located within the buffer zone of an area of ​​interest 32 shifted to the positions 44 highlighted here.

[0048] Figure 6 shows an illustration of a combined overlap map, i.e., a superimposition of the overlap maps according to Figure 4 and 5 and also with reference to the light spot position 30. The separate folding of the actual area of ​​interest 32 and the buffer zone with subsequent combination serves primarily for illustrative purposes.

[0049] A joint folding is also possible, whereby different weights can be taken into account in the folding to reflect the respective importance of, for example, the buffer zone.

[0050] In the combined overlap map of the Figure 6Is there a suitable area 46 with good positions where neither the area of ​​interest 32 nor its buffer zone would enclose a defective light-receiving element 28? In the unsuitable areas 48, however, at least one defective light-receiving element 28 would fall within the area of ​​interest 32. At positions in the prohibited area 50, the area of ​​interest 32 would simply no longer lie completely on the light receiver 26. At positions in the boundary areas 52, the buffer zone would either contain at least one defective light-receiving element 28 or would not fit completely on the light receiver 26.

[0051] Figure 7 shows the result of an evaluation of the combined overlap map according to Figure 6, which can be interpreted as a map with a defect scale. The defect scale is light for prohibited or unsuitable positions and increasingly dark for particularly suitable positions. In addition to the ratings, which already include Figure 6 The defect measurement can include a distance to the original light spot position 30 and / or take into account whether the position lies within the specified target area 36 or not. Further evaluation criteria introduced by users are conceivable.

[0052] In the example of the Figure 7There are numerous suitable (dark) positions where no defective light-receiving elements 28 would be located, either in the area of ​​interest 32 or in its buffer zone. The new light spot position 30a is selected from these because it requires the smallest displacement. If the light spot position can be moved arbitrarily, a distance-dependent weighting can be omitted, and there is then corresponding freedom of choice for the new position of the area of ​​interest 32.

[0053] In Figure 7In the upper left corner, a dark rectangle appears with particularly suitable positions where the defect size is improved due to the specified target area 36. Equivalently, all positions outside the target area 36 are penalized accordingly in terms of the defect size. However, due to the large distance to the original light spot position 30, the comparatively small advantage of target area 36 over the nearby new light spot position 30a does not materialize.

[0054] Figure 8 The result is shown again in the original defect map, which is located in the bottom right corner. Figure 2The area of ​​interest 32 is shifted to the new light spot position 30a and contains no defective light-receiving elements 28 at this position. As explained above, the shift of the light spot is achieved by mechanical manipulation, particularly during manufacturing. The area of ​​interest 32 can still be adjusted to the actual light spot position 30a or even definitively fixed at this point to compensate for any tolerances.

[0055] Figure 9Finally, a comparative representation of the increased yield of sensors 10 thanks to the shift of the area of ​​interest 32 while avoiding defective light-receiving elements 28 is shown. The probability of a defect per light-receiving element 28 is plotted on the x-axis, and the yield on the y-axis. The different measurement curves correspond, from bottom to top, to an allowable shift of zero to four light-receiving elements 28 on the light receiver 26. The lowest measurement curve therefore represents the state of the art without an optimized position of the area of ​​interest 32. With the highest measurement curve, allowing a shift of four, there are already virtually no failures; this can be further improved with even more flexible shifts.

[0056] In this example, it is assumed that the area of ​​interest 32 comprises a total of 18 light-receiving elements 28. The probability that none of these are defective when selected randomly is only 88-92%. If a defect were tolerated, this probability would increase to over 99%, but the defective light-receiving element 28 in the area of ​​interest 32 would already affect the performance of the sensor 10. The invention makes it possible almost always to select the area of ​​interest 32 without a defective light-receiving element 28, and therefore no loss in yield or performance has to be accepted.

[0057] The described method already incorporates a buffer zone to accommodate manufacturing tolerances. It is also possible to account for other effects using the method described above. Figures 2 to 8The described procedure takes into account factors such as temperature changes, shading, adherence to predefined (cross-eyed) angles, and the like.

Claims

1. A method of configuring an optoelectronic sensor (10) that has a light receiver (26) having a plurality of light reception elements (28) and that has a control and evaluation unit (34), wherein a region of interest (32, ROI) having only some of the light reception elements (28) is selected and the control and, after the configuration, the evaluation unit (34) only evaluates information from light reception elements (28) of the region of interest (32) and wherein defective light reception elements (28) are determined and the region of interest (32) is selected such that the region of interest (32) comprises at most a maximum number of defective light reception elements (28), characterized in that a defect map of the light receiver (26) having highlighted positions of the defective light reception elements (28a-b) is generated from the determined defective light reception elements (28a-b); and in that the defect map is folded together with an ROI map corresponding to a shape of the region of interest (32) to form an overlap map.

2. A method in accordance with claim 1, wherein the selection of the region of interest (32) takes place as part of the manufacture of the sensor (10).

3. A method in accordance with claim 1 or claim 2, wherein the defective light reception elements (28) are determined during the detection of a reference object by the sensor (10).

4. A method in accordance with any one of the preceding claims, wherein the sensor (10) has a light transmitter (12) whose transmitted light generates a light spot on the light receiver (26) and wherein the region of interest (32) is selected such that the light spot is incident on light reception elements (28) of the region of interest (32).

5. A method in accordance with claim 4, wherein a light spot position (30) on the light receiver (26) is displaced in that the light transmitter (12), a transmission optics (14) of the light transmitter (12), the light receiver (26), and / or a reception optics (24) of the light receiver (26) are moved.

6. A method in accordance with any one of the preceding claims, wherein the ROI map has a buffer zone about the region of interest (32).

7. A method in accordance with any one of the preceding claims, wherein positions of the overlap map are evaluated by a defect measure, wherein the defect measure adopts a value depending on the overlap of the ROI map with defective light reception elements (28a-b).

8. A method in accordance with claim 7, wherein the region of interest (32) is selected such that the defect measure for the positions of the light reception elements (28) of the region of interest (32) does not exceed a maximum defect measure.

9. A method in accordance with claim 7 or claim 8, wherein the defect measure is reduced at positions within a predefined region (36) of the light receiver (26).

10. A method in accordance with any one of the claims 7 to 9, wherein the defect measure has a portion that depends on a distance of the respective position evaluated by the defect measure from an origin position of the respective (32), in particular from an original light spot position (30).

11. A method in accordance with any one of the claims 1 to 6, wherein the region of interest (32) is selected such that the region of interest (32) does not comprise any defective light reception elements (28).

12. A method in accordance with any one of the preceding claims, wherein the light reception elements (28) are configured as Geiger mode avalanche photodiode elements.

13. A method in accordance with any one of the preceding claims, wherein the sensor (10) has a plurality of time of flight measurement units that are changeably connectable to light reception elements (28) and wherein the selection of a region of interest (32) takes place by the connection between the light reception elements (28) and the time of flight measurement units.