MXPA98002452A - Inspection of the pressure and the filling level of a container using multidimension images - Google Patents

Abstract

The present invention relates to a container inspection system for inspecting a moving container (116) including a radiation source (102) positioned to direct the radiation towards the moving container. A radiation detector (104) is positioned to receive a portion of the radiation from the radiation source that is not absorbed or blocked by the moving vessel and to generate electrical signals in response to it. The processing circuitry (106) produces multidimensional image data for the moving vessel based on the electrical signals generated by the radiation detector, and compares at least a portion of the multidimensional image data with a corresponding portion of the data multidimensional image for a standard container. Subsequently, the processing circuit determines, based on a result of the comparison, one or more characteristics that include the filling level of the container, regardless of whether the container is overfilled, or has insufficient filling, or the container is properly pressurized

Description

INSPECTION OF THE PRESSURE AND THE FILLING LEVEL OF A CONTAINER USING MULTIDIMENSIONAL IMAGES BACKGROUND OF THE INVENTION In the process of filling containers, for example in canning and bottling lines, it is usually useful to monitor the characteristics of the containers that are being filled. For example, the levels at which containers are being filled can be monitored for quality control purposes. It is known to use a radiation source and a detector to determine the filling level of a container. For example, Schiessl et al., U.S. Patent No. 4,481,595, describes a system that makes the containers pass through a beam of gamma radiation projected from a beam source to a detector. As the vessel passes through the beam, the system counts the pulses of radiation received by the detector. Once the entire vessel has passed through the beam, the system determines the average rate at which the radiation pulses were received by the detector and compares this rate with a reference rate. Based on this comparison, the system produces a signal indicative of whether the material inside the container is at a high enough level to attenuate the beam. The system can be configured to detect insufficient filling conditions by orienting the source and the detector so that the detected pulses pass through the vessel at a lower level than the expected fill level. Similarly, the system can be configured to detect overfilling conditions by orienting the source and the detector so that the detected pulses pass through the container at a level above the expected fill level, SUMMARY OF THE INVENTION This invention includes a container inspection system that produces a multidimensional image of each container to be inspected. The system then analyzes the image to provide real-time monitoring of the characteristics such as filling level of the product, presence and proper placement of caps, pressure inside the container, density of the foam at the top and leakage in the moving containers. at typical process rates on a conveyor of a container filling process. In the sense used herein the term "container" refers to cans, bottles and other containers whose intended content is generally known. The system provides an accurate measurement of the conveyor speed in the order of 2400 containers per minute, and is capable of inspecting containers made from a wide range of materials, including metals, plastic, glass and aluminum foil. If the system determines that a container is not properly filled, or has inadequate pressure or any other defect, the system automatically initiates appropriate actions in order to reject the container from the filling line and / or adjust the filling operation. The system keeps a complete record of all the rejections and their causes, an operator of the system can use this diagnostic data for the maintenance and improvement of the efficiency of the process. The system provides considerable advantages over the prior art systems which provided only gross indications of "pass / fail" or "insufficient filling / overfilling". For example, the system uses multidimensional information about the containers to provide fill level measurements that have an accuracy within 0.5 mm in a range of inspection speeds. This high level of accuracy allows having narrower fill level thresholds and, therefore, reducing the number of false rejections, which in turn improves the efficiency of the inspection process. The system uses a radiation source, for example an X-ray energy source with a multi-element linear detector to inspect the filled containers moving on a conveyor line. As the container moves on the conveyor line, it passes between the radiation source and the detector arrangement so that the radiation produced by the radiation source passes through the container before it is detected by the detector array. Due to differences in path length and radiation absorption coefficients, radiation is absorbed differently by the container, container lid, contents of container and air or any other material above the container. content of the container. These differences in absorption are measured as changes in the intensity of radiation received by the detector array. When the conveyor is oriented to move the containers in a horizontal direction, the radiation source and detector arrangement are placed to define a vertical plane between the source and the detector, and are oriented so that the plane is perpendicular to the direction of the movement of the transporter. Consequently, at any given time, the radiation received by the detector array corresponds to a vertical interval or vertical slice of a container. By repeatedly receiving and storing the detector array data as the conveyor moves to the vessel, the system produces a multi-dimensional image of the vessel, wherein the resolution of the image is controlled by the number of elements in the detector array and the frequency to which the data is received and stored. Subsequently, the system processes the image data to monitor the characteristics such as fill level and pressurization and detects conditions such as insufficient filling, excessive filling, low pressure, high pressure, missing or damaged caps and beaten or bulging containers. When determining the level of filling, the system can determine the presence of foam, by determining the density of the foam and the level (amount) of liquid attributable to the foam and adding the quantity to the apparent level of filling (quantity). The system can also monitor conditions such as container wall thickness. The system provides several advantages over the prior art. In particular, the system monitors excessive filling, insufficient filling, actual filling level, containers with low pressure, missing lids, dented containers, wall thickness of the container and foam characteristics. Significantly, the system performs all these operations simultaneously using a single sensor. The system accurately determines the level of filling and other characteristics, even in inspection areas where there is significant agitation of the contents of the containers (ie the performance of the system is not affected by the movement of the contents of the container). The system compensates for this movement by collecting information about the presence of the liquid in a relatively large area of the container and combining the information to determine the level of filling. This allows the system to be placed, for example, on or immediately after a conveyor curve or immediately after the containers have been turned over. The system monitors the low pressure (leaks) of the containers without manipulating them. In contrast, in the prior art, leaking containers are detected by inverting the containers, allowing the liquid to drain and subsequently detecting the low pressure of the container using an insufficient filling detector. This required means to invert the containers and also required that the leaks be large enough to allow sufficient liquid to be drained from the container during the inspection process. The system is easily calibrated by passing a standard container or calibrator through the system and producing a standard image that includes all pertinent information about the desired characteristics of the containers to be inspected. The system automatically adjusts the height of the container and can therefore adjust the changes in the size of the container during production, without recalibration. For example, the system may include a motorized shelf that automatically places the unit at a pre-set inspection point. The system is also relatively insensitive to variations in the position of the container due to wear of the conveyor or other factors. Wear of the conveyor, for example, can cause one or more containers to be placed lower than others. When the system detects a container placed imperfectly in these cases, the system automatically adjusts the inspection area to take into account the change in the position of the container. To reject unacceptable containers, the system uses an intelligent rejection system. The sensors monitor the rejector's performance to verify proper rejects and gather information about wear factors and other factors. This information is used to compensate for the effects of wear and allows early diagnosis and correction of problems. A dual rejection can be used to reject two successive vessels and to provide redundancy if a rejector fails. The ability of the system to accurately measure the level of failure can also be used to monitor and adjust the operation of the filler. By constantly adjusting the filling valves, the system optimizes the appliance's performance and reduces waste. Other features and advantages of the invention will be more apparent from the following description of the preferred embodiments, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a block diagram of a container inspection system. Figure 2 is a front view of one embodiment of the container inspection system of Figure 1. Figure 3 is a rear view of the container inspection system of Figure 2. Figure 4 is a side view of the inspection system of containers of Figure 2, showing the side through which the containers enter the container inspection system. Figure 5 is a side view of the container inspection system of Figure 2, showing the side through which the containers exit the container inspection system. Figure 6 is a plan view of the orientation of the X-ray source and the X-ray detector of the container inspection system of Figure 2, with respect to a container. Figure 7 is a block diagram of the detector and the container inspection system controller of Figure 2. Figures 8 to 12 are flow charts of procedures instrumented by the container inspection system controller of Figure 2. Figure 13 is a graphical view of the image data produced by the container inspection system of Figure 2. Figures 14 to 16 are block diagrams showing the placement of a container inspection system.

DESCRIPTION OF THE PREFERRED MODALITIES Referring to Figure 1, a container inspection system 100 includes an X-ray source 102, a diode detector array, linear, multi-element 104, a controller 106 and a rejector 108. The source X-ray detector 102 is configured to produce a vertically flat x-ray beam 110, which is received by the detector array 104. The beam 110 is perpendicular to a direction of movement 112 of a conveyor 114. The detector array 104 includes 32 elements of diode, each of which provides the controller 106 with an analog signal corresponding to the incident of X-ray radiation on the diode. As a container 116 (e.g., a soda can) approaches the x-ray beam 110, the container interrupts a light beam 118 between the light source 120 and an optical container activator 122 which causes the activator 122 of container sends a signal to controller 106. Controller 106 responds by periodically storing the analog signals received from detector array 104. At any time, a sweep of signals produced by detector array 104 corresponds to a dimensional X-ray image ( vertical) of the container 116 as it passes in front of the arrangement. Successive sweeps are made as the movement of the conveyor 114 causes the container 116 to traverse the face of the detector array 104 in the horizontal direction. The controller 106 synchronizes the successive sweeps of the detector array with the movement of the vessel 116, simultaneously monitoring the output of an encoder 124 which is mechanically linked to the conveyor 114. The encoder 124 produces a series of pulses where each corresponds to a portion of a rotation of a conveyor drive shaft 114. The controller 106 counts these pulses to monitor the position of the containers 116. This allows the controller 106 to control the horizontal scan rate based on the instantaneous velocity of the container, so that each Vertical sweeping starts at a fixed horizontal distance (independent of speed) with respect to the leading edge of the container. In this form, the controller 106 stores in the memory an exact two-dimensional image of the X-ray absorption characteristics of the container 116, as it passes in front of the detector arrangement 104. Once the container 116 has completely passed through the through the X-ray beam 110, the controller 106 processes the image data to determine whether the container 116 is inadequately filled or has defects. If so, the controller 106 activates the rejector 108 and this rejects the defective container 116 from the conveyor 114. In addition to the container activator 122, the system includes a reject trigger 126 which produces a signal in response to the interruption of an optical beam 128 produced by a light source 130. The rejection actuator 126 is used to verify the position of a container 116 before rejecting the container 116. The operation of the system is based on the assumption that there is no slip (ie, that a position of the container in the conveyor does not change). The use of the reject actuator 126 allows some displacement of the containers as long as one container does not move to a full position of another container (i.e., the diameter of the container) between the location of the container activator 122 and the activator 126 of the container. rejector As desired, the system may also include other optical sensors, including for example, a dropped vessel sensor (not shown) and an optical sensor (not shown) that monitors the entrance of a reject chute 132 for the passage of a vessel. 116. With reference also to Figures 2 to 5, the inspection system 100 includes a cabinet 200 mounted on an adjustable shelf 202. The cabinet 200 contains the X-ray source 102, the controller 106 and the electronic support components. The detector array 104 is mounted on an adjustable tunnel unit 204 which is also connected to the cabinet 200. The adjustable tunnel unit also supports the optical container activator 122. Accordingly, the system can be adjusted for a change in the container size by adjusting the vertical position of the tunnel unit 204. The position of the tunnel unit 204 is controlled and monitored by the controller 106. Typically, the controller 106 can adjust the height of the tunnel unit within a range of ten inches, which allows the size of the containers to vary between, for example seven ounce cans and forty ounce bottles. The system could also be adjusted by automatic or mechanical adjustment of the vertical position of the shelf 202. A user interface 206, including a video screen 208 and input keys 210, is provided in front of the cabinet 200. A warning light 212 indicates that the system is working. In addition to providing support for the source, the detector, the electronic components and the user interface, the cabinet 200 provides protection to the operator of the system against exposure to X-ray radiation. The supporting electronic components include amplifiers that amplify the signals analogs produced by the detector array 104 and power sources for the system and the X-ray source. The system also includes a slot unit 402 (see Figure 4) to collimate the X-ray beam produced by the X-ray source 102. The alignment of the source 102 causes the beam 110 and the detector 104 to be maintained through the connection of the source 102 and the detector 104, with the cabinet 200. Accordingly, as illustrated in Figure 4, the system can be easily installed by positioning it so that the tunnel unit 204 is mounted on the conveyor 114 and aligned horizontally therewith. The system only needs to be roughly aligned in the vertical direction since, as mentioned, the vertical position of the tunnel unit is automatically adjusted during an initialization procedure that allows rapid change from one container size to another. Also shown in Figure 3 is the rejector 108, which includes a pair of air-driven rams 302. Each ram 302 includes an air pressure cylinder and a solenoid, and is controlled independently of the controller 106. The use of two rejectors allows the rejection of the containers 116 at conveyor speeds of up to 2400 containers per minute, alternating the duty cycle of each ram 302, as required by the rejection conditions of the system. The sensors (not shown) associated with each ram 302, monitor the condition of the ram providing an indication of the time the ram comes out of its rest position and the time it returns to it. An optical sensor (not shown), which is mounted on the reject ramp 132 (Figure 1) verifies that a desired container 116 has indeed been rejected and detects any unwanted rejection. The operation of the rejector 108 is completely automatic, the system tracks the position of a container 116 to be rejected, rejects the defective container, verifies the rejection and monitors the condition of the reject ram 302. The X-ray source 102 provides a beam continuous X-ray at 40-70 kV and 0.01 to 0.08 mA (ie, 0.4-5.6 W). The energy level is adjustable for different types of containers (eg, aluminum versus stainless steel) by jumpers on a control board within the cabinet 200. The energy level can also be adjusted by the controller 106. Typically, the energy level it is thickly adjusted based on the type of container to be inspected and finely tuned later to provide adequate contrast. The controller 106 monitors the operating energy of the X-ray source. In the embodiment described, the X-ray source is supplied by Lorad Division, ThermoTrex Corporation, Danbury, Connecticut. The use of a continuous source eliminates the timing problems associated with pulse sources. As illustrated in Figure 6, the X-ray source is a one-millimeter 600-point source that is collimated through a slot unit 402 to produce the X-ray beam 110. The slot 602 of the Slot 402 is one millimeter wide and fifteen millimeters high. To increase the resolution of the system, an X-ray source having a smaller point source can be used. As illustrated in Figure 6, the X-ray beam 110 is oriented so that it passes only through an upper portion 606 of the container 116. As mentioned below, the X-ray absorption characteristics of this region of the container They include all the necessary information to determine if the container is defective or not. Of course, when desired or necessary, the X-ray beam 110 could be oriented to produce an image of the entire container 116. As illustrated in Figure 7, the detector array 104 includes two arrays 700 of 16 elements. The photosensor surface of each diode 702 of the arrays is two millimeters wide and one millimeter high, and diodes 702 improve sensitivity to ultraviolet radiation. Although each diode is 1 millimeter high, the detector array provides vertical resolution in the order of 0.5 millimeters. This resolution increase occurs because the beam is projected at an angle through the container and because a portion of the container placed between the vertical centers of two adjacent diodes 702 will affect the two diodes and, therefore, may be identified. by variations in the signals produced by the two diodes. A cesium / iodine segmented crystal 704 scintillator that converts the incident x-ray radiation to ultraviolet radiation, overlies each array 700. In the embodiment described, the arrays 700 are supplied by Photonics Corporation. A phosphor screen could be replaced by the glass scintillator 704. However, the scintillator can be preferred since it provides a faster response, the use of a phosphor screen can also make the image blurry. In addition, a non-segmented phosphor screen would tend to increase the interference between the diodes. The analog signal produced by each diode 702 is amplified by a dedicated amplifier in an amplifier board 706. The amplified signals are supplied to a multiplexer 708 of 32 to 16, which is controlled by a signal coming from the controller 106. The signals produced by the multiplexer are supplied to an analog input channel 710 of sixteen bits of the controller 106. Each bit of the analog input channel is converted into a digital value with twelve resolution bits. Typically, the controller 106 is implemented using an 80486 processor available from Intel Corporation.

Referring to Figure 8, the controller 106 controls the system 100 according to a procedure 800. To initiate the operation of the system, a user selects the initialization using the keyboard 210 of the user interface 206 (see Figure 1). In response, the controller 106 implements an initialization and calibration routine 802. After initialization, the controller operates the system according to a detection or acquisition routine 804 which detects a container 116 and acquires the data for the container 116. Al To complete that routine, the controller 106 operates the system according to an analysis routine 806 to determine if the container is defective. If the controller 106 determines that the container 116 is defective, the controller operates the system according to a rejection routine 808. It is important to note that the system can operate simultaneously according to the routines of detection, acquisition and rejection. For example, the system could operate to reject a first container at the same time that it is analyzing the data for a second container and acquiring data for a third container. In the described embodiment, the controller 106 is fast enough to complete the analysis of the data for one container while acquiring the data for another container. Accordingly, the controller 106 includes two data buffers, each of which is of sufficient size to store the data of a container. Referring to Figure 9, the controller 106 starts the initialization and calibration routine 802 by determining the gain and offset of each diode 702 of the detector array 104 (step 900). As is well known, the voltage produced by a diode 702 corresponds to the diode bypass voltage plus the product of the incident x-ray radiation on the diode and the diode gain: V = incident * gain + offset. Accordingly, when the gain and deflection of a diode is known, the incident x-ray radiation of the diode can be determined from the voltage produced by the diode. Since the gains and variances vary from diode to diode, the controller 106 determines and stores the gain and deviation of each diode, and uses these values when the signals produced by the diodes are processed. The controller 106 determines the deviation of each diode by measuring the voltage produced by each diode when the x-ray source 102 becomes incapacitated so that there is no incident x-ray radiation on the diode: V = gain * 0 + deviation = deviation. Once the deviations are known, the processor determines the gain of each diode by subtracting the deviation of the diode from the voltage produced by the diode when the x-ray source 102 is turned on and no vessel interrupts the x-ray beam 110: V - deviation = gain * 1 = gain, where the incident x-ray radiation is normalized so that a value of 1 corresponds to an uninterrupted beam and a value of 0 corresponds to an incident radiation. Subsequently, the controller 106 controls the adjustable shelf 202 to elevate the system to its highest vertical position (step 902) and prompts the system operator (through the user interface 206) to place a test container on the conveyor 114 Subsequently, the controller 106 monitors the signals produced by the detector array 104 to determine if the vertical position of the system is correct (step 904). In the described embodiment, the correct vertical position is defined as the position in the incident x-ray radiation over the fifth diode 702 from the top of the detector array 104 is less than or equal to 70% of a complete beam (i.e. , the test container blocks at least 30% of the x-ray radiation directed to that diode). If the vertical position is not correct, the controller 106 instructs the adjustable rack 202 to lower the system in an increment (step 906) and check the position again. Once the vertical position of the system is correct, the controller prompts the operator to place the test container on the conveyor and measures the diameter of the test container (step 908). In the described embodiment, the controller 106 measures the diameter of the container relative to the speed of the conveyor 114, counting the number of pulses produced by the encoder 124 from the moment when the test container interrupts the optical beam 118 and activates the container activator 122 until such time as the test container passes outside the optical beam 118 and deactivates the container activator 122. At the same time, the controller 106 determines the relationship between the encoder pulses and the horizontal distance, counting the number of encoder pulses that occur between activation of the container activator 122 by the test container and activation of the rejection actuator 126 by the test container. Because the distance between these triggers is known, the pulse distance of the encoder can be determined by dividing the known distance between the pulse count. Subsequently, the controller 106 identifies the edge and center of the test container (step 910). Once the test vessel interrupts the optical beam 118, the controller 106 stores the values of the signals produced by each diode 702 by successive horizontal increments (typically of the order of every two encoder pulses). Based on other values, the controller 106 identifies the edge of the test container as corresponding to the first set of signals where a portion of the x-ray radiation incident on the fifth diode 702, which comes from the top of the diode array It is interrupted by the test container. After identifying the edge of the test container, the controller 106 identifies the center of the test container corresponding to the set of signals separated from the edge, in the middle of the number of pulses of the encoder corresponding to the diameter of the container. Once the edge and center of the test container have been identified, the controller 106 identifies the values that correspond to the regions of the image that are of particular interest. As illustrated in Figure 13, in the embodiment described, where the containers are soda cans, the image data includes 64 columns of data, each of which includes 32 entries (corresponding to the 32 diodes of the diode array). The leading edge of the can is presented in column 12 and the center is presented in column 38. There are two regions of interest. The first region 1300, which corresponds to the upper part of the can and is used to determine whether it is adequately pressurized, includes columns 35 to 41 of rows 24 to 27. The second region 1302 is used to measure the level of liquid in the can and includes the columns 23 to 56 of rows 12 to 23. Finally, using the values corresponding to the regions of interest, the controller 106 generates threshold values for each region of interest (step 914). For the first region 1300, the tab on the top of the can is expected to be placed in the center of the region. Accordingly, the controller 106 multiplies the values corresponding to the rows 26 and 27 by a positive weighting factor, multiplies the values corresponding to the rows 24 and 25 by a negative weighting factor, and sum all the values to produce the threshold value. For the second region 1302, the controller 106 adds all the values to produce the threshold value. By summing all the values, the controller 106 generates a measurement of the x-ray absorption properties of the entire second region 1302. This is extremely significant since it results in the system's ability to measure the filling level, so that be insensitive to the agitation of the contents of the container. In the prior art, fill level sensors typically had to be placed at least 15-30 feet downstream of a stirring source, for example a conveyor curve or a filling station to allow the content of the containers will settle before the analysis. In contrast, the inspection system of the container 100 may be placed in a curve or immediately after a source of agitation without deleterious results. Referring to Figure 10, the controller 106 initiates the detection and acquisition routine 804 determining whether the container activator 122 has detected a container (step 1000). If so, the controller 106 initializes a delay timer to a value corresponding to the number of encoder pulses that are expected to occur before the leading edge of the container is properly positioned and initializes a zero count (step 1002). Subsequently, the controller monitors the pulses of the encoder until the delay timer has expired (step 1004). After the delay timer expires, the controller 106 stores the measurement values from the array of the diode and increases the measurement count (step 1006). As discussed above, the measurement values are generated by modifying the corresponding number of the voltage of each diode by the deviation and gain of that diode. If 64 measurements have not been taken (step 1008), the controller waits for the occurrence of an adequate number of encoder pulses and repeats the storage and increment steps (step 1006). Once the 64 measurements have already been taken, the controller 106 starts the analysis routine 806 and simultaneously starts the detection and acquisition routine for the next container 116. Referring to Figure 11, the controller 106 starts the routine of analysis 806 identifying the position of the upper part of the container within the measured data (step 1100). By allowing the position of the upper part of the container to vary, the controller 106 takes into account the variations in the height of the conveyor that could result, for example, from components worn unevenly on the conveyor. Once the back of the container is identified, the controller determines the regions of interest for the container (step 1102). As mentioned before, the upper part of the test container is placed in row 28 (ie, in the fourth diode from the top) and the first region 1300 is defined in rows 24 to 27. In this way, if the upper part of the container were identified in row 29, the first region 1300 would be defined in the rows 25 to 28. After identifying the regions of interest, the controller 106 generates numbers of each region of interest using the procedure described above for general thresholds (step 1104). These numbers are then compared to the thresholds (step 1106). If one of the numbers varies from the corresponding threshold by a predetermined percentage, the controller 106 determines that the container must be rejected (step 1108). When the controller 106 determines that a container must be rejected, the controller executes the reject routine 808. Referring to Figure 12, the controller 106 initiates the rejection routine 808 expecting the container to interrupt the optical beam 128 of the reject trigger. 126 (step 1200). When this occurs, the controller 106 knows the exact position of the container and responds by initializing a counter that counts the pulses from the encoder 124. (step 1202). The controller 106 then counts the pulses until the count indicates that the container is positioned so as to activate a reject ram 302 (1204). Subsequently, the controller activates the reject ram 302. As already mentioned, the controller 106 activates the rejection earrings 302 in an alternating fashion. In this way, the pulse count that is indicative of the proper position of the container will vary based on what the rejection rams 302 will activate. It is also important to note that, due to the speed of the conveyor 104 with respect to the speed of the rejecting rams 302, a reject ram 302 will typically be activated before the vessel is placed in front of the reject ram, and a return signal from the reject ram Your resting position may be emitted before the container reaches the ram. The controller 106 modifies the pulse count corresponding to the proper position of the vessel, based on the feedback signals received from the rejection rams. This allows the controller 106 to account for changes in the operating characteristics of the rejection rams over time. As illustrated in Figure 14, two or more inspection systems 100 may be employed to provide fail-safe operation. When two inspection systems 100 are employed, the systems are sequentially positioned along the conveyor 114 and share a common rejector 1400 that is placed downstream of the systems relative to the direction of movement 112 of the conveyor. With this arrangement, each system 100 inspects all containers and rejects those found defective. Each system 100 monitors the signals sent to the rejector 1400 by another system 100 and compares the signals with those generated to verify the proper operation of the system and detect system failures. As illustrated in Figure 15, the container inspection system 100 is typically placed downstream, relative to the direction of movement 112 of the conveyor 114, of a filler 1500 that fills the containers and of a crimper 1502 that seals the containers. The feedback paths 1504 from the system 100 to the filler 1500 and the engargoladora 1502 allow the automatic adjustment of these components. For example, the filler 1500 can adjust a fill valve in response to information that comes from the system 100 indicating that the fill valve is not functioning properly. In a similar way, the engargoladora 1502 can make adjustments in response to information indicative of improperly engargolados containers. Finally, as illustrated in Figure 16, the ability of the container inspection system 100 to accurately determine the level of filling of the containers allows the system to be placed immediately downstream of a curve 1600 on the conveyor 114 Other modalities fall within the following claims. For example, to improve resolution, the x-ray beam 110 could be focused using, for example, a tungsten honeycomb structure or the number of elements in the detector array could be increased. Similarly, a detector array having a higher element density in a particular region of interest could be employed. In addition, the source of x-rays could be replaced with a source of gamma radiation. However, x-ray radiation is preferred over gamma radiation because, for a particular power level, x-ray radiation provides more information. While the system described above is primarily configured to inspect cans that are expected to have identical characteristics nearby, they could also be used to inspect bottles or other containers where the wall thickness of the container varies from container to container or even within the same container. When inspecting these variable containers, the system would determine the wall thickness of each container and take into account this for the effects of variation in thickness. Also, unlike cans, filled bottles typically include a large top space where varying levels of foam are formed. To determine if a bottle is properly filled, the system would detect a level of foam in the bottle and, based on the density of the foam, would modify the liquid level measured accordingly. In an approach to analyzing the foam, the controller 106 searches for positive gradients in the x-ray attenuation between the horizontal rows of a region of interest in the image data. The controller uses the location of these gradients to determine the relative position of the foam-liquid boundary. Once the boundary is located, the controller determines the volume of the foam based on the known geometry of the container and assumes that the foam fills the entire volume of the container above the foam-liquid boundary. The controller determines the density of the foam by comparing the absorption measurements that come from the detector elements, immediately above and below the limit, where the measurement from below the limit corresponds to the absorption by liquid and the measurement from above the limit corresponds to the absorption by foam. Subsequently, the controller determines the amount of liquid in the foam, multiplying the volume of the foam by the density of foam. Finally, the controller adjusts the measured fill level according to this amount. When appropriate, an air / foam limit would also be detected, and its position could be used to determine the volume of foam in the container. When examining a glass container, the controller estimates the thickness of the container walls by measuring the attenuation gradient along the vertical edges of the container. The controller can use the thickness of the glass as a first order of correction for the volume of the container at both the filling level and the foam measurement. In another approach to analyzing the image data, the image data for the regions of interest for a large number of containers (for example from 100 to 1000) could be used to form a neural network. Subsequently, the containers could be inspected by applying their image data to the neural network.

Claims (38)

1. CLAIMS: 1. A system for determining the level of filling or pressurization of a moving vessel, comprising: a radiation source positioned to direct the radiation in the moving vessel; a radiation detector positioned to receive a portion of the radiation from the radiation source that is not absorbed or blocked by the moving vessel and to generate electrical signals in response to it; and processing circuitry that functions to: produce multidimensional image data for the moving vessel based on the electrical signals generated by the radiation detector; comparing at least a first portion of the multidimensional image data with a corresponding portion of the multidimensional image data for a standard vessel and determining, based on a result of the comparison, one or more characteristics of the vessel from the set of characteristics that include the filling level of the container, if the container is overfilled, if the container is not sufficiently filled, if it is adequately pressurized or if it is properly sealed. The system according to claim 1, wherein the processing circuitry functions to determine that the movement vessels differ from the standard vessels, when the portion of the multidimensional image data for the moving vessel differs from the portion of the data. multidimensional image for the standard container in more than a predetermined amount. 3. The system according to claim 2, wherein each of the first portion and the corresponding portion includes a plurality of elements and wherein the processing circuitry functions to combine values associated with each of the plurality of elements of the first portion, so as to producing a first composite value, to combine values associated with each of the plurality of elements of the corresponding portion, in order to produce a second composite value and compare the first and second composite values to determine the characteristics of the container. The system according to claim 3, wherein the first portion includes multidimensional image data corresponding to a region above and below the expected fill level in the moving vessel, and wherein the processing circuitry functions to determine a current filling level of the container, as the characteristic of the container. The system according to claim 4, further comprising a rejector, wherein the processing circuitry functions to activate the rejector when the filling level of the container differs from a filling level of the standard container by more than a predetermined amount. The system according to claim 3, wherein the first portion includes multidimensional image data corresponding to a region that includes an upper surface of the container and wherein the processing circuitry operates to determine whether the moving container is adequately pressurized, as the characteristic of the container. 7. The system according to claim 1, wherein the radiation detector comprises linear, one-dimensional detector array, which operates to produce electrical signals corresponding to a one-dimensional segment of the moving vessel, and the processing circuitry operates to produce the multidimensional image data. assembling sets of electrical signals that correspond to one-dimensional segments of the moving vessel. The system according to claim 7, wherein: the moving container moves in a horizontal direction; the radiation detector comprises a linear, vertical, one-dimensional detector array, and the radiation detector operates to generate the sets of electrical signals as the container moves beyond the detector array. 9. The system according to claim 1, wherein the radiation source functions to direct radiation in only a portion of the moving vessel. The system according to claim 1, wherein: the moving container comprises a can whose contents are sealed therein, the first portion of the multidimensional image data corresponds to a region on the top of the can, and the processing circuitry functions to determine that the moving vessel is not properly pressurized when the multidimensional image data for the region on the top of the can differ from the multidimensional image data for the region at the top of the standard vessel . 11. The system according to claim 1, wherein the radiation source comprises a continuous X-ray source. The system according to claim 1, further comprising a mechanism that functions to automatically adjust the vertical position of the radiation source and the radiation detector based on the height of the moving vessel. The system according to claim 1, wherein the moving container is placed on a conveyor, the system further comprises a rejector that functions to remove the moving container from the conveyor, when the processing circuitry determines that the moving container differs from the standard container. The system according to claim 13, wherein the rejector comprises an air driven ram. 15. The system according to claim 14, wherein the rejector comprises a pair of air-driven rams, operating in an alternating manner. The system according to claim 14, wherein the rejector operates to provide the processing circuitry with information about the movement of the air-driven ram. 17. The system according to claim 7, wherein the radiation detector comprises an array of photodiodes. 18. A method for inspecting a moving vessel, comprising: directing radiation in the moving vessel; receive a portion of the radiation that is not absorbed or blocked by the moving vessel; producing multidimensional image data for the moving vessel based on the received radiation that was not absorbed or blocked by the moving vessel; comparing at least a first portion of the multidimensional image data with a corresponding portion of the multidimensional image data for a standard container; and determining a filling level of the moving vessel based on a result of the comparison step. A method according to claim 18, further comprising determining that the moving vessel differs from the standard vessel when the portion of the multidimensional image data for the moving vessel differs from the portion of the multidimensional image data for the standard vessel , in more than a predetermined amount. A method according to claim 18, further comprising: directing radiation in the standard container so that a portion of the radiation is absorbed or blocked by the standard container and a portion of the radiation is neither absorbed nor blocked by the standard container; receive the portion of the radiation that is not absorbed or blocked by the standard container; and producing the multidimensional image data for the standard container, based on the received radiation that was not absorbed or blocked by the standard container. 21. A method according to claim 18, wherein: the step of receiving radiation comprises receiving radiation as a series of one-dimensional segments; and the step of producing the multidimensional image data comprises producing a set of one-dimensional image data for each of the one-dimensional segments of the received radiation and assembling the one-dimensional image data set to produce the multidimensional image data. The method according to claim 21, further comprising moving the moving container in a horizontal direction, wherein the step of receiving the radiation comprises receiving radiation as a series of one-dimensional vertical segments, as the moving container moves. beyond a vertical detector. The method according to claim 18, wherein the step of directing radiation comprises directing radiation to only a top portion of the moving vessel. The method according to claim 18, wherein: the moving container comprises a can, the first portion of the multidimensional image data corresponds to a region at the top of the can, and the step of determining comprises determining that the moving vessel is not properly pressurized when the multidimensional image data for the region on the top of the can differ from the multidimensional image data for the region at the top of the vessel standard. The method according to claim 18, wherein the radiation comprises a continuous beam of X-ray radiation. 26. The method according to claim 4, wherein the processing circuitry functions to take the foam in the container into account, when the actual filling level of the container is determined. 27. The system according to claim 26, wherein the processing circuitry operates to determine an apparent fill level of the container, to determine a liquid level attributable to the foam in the container, and to combine the apparent fill level and the attributable level to the foam in order to determine the actual filling level. The system according to claim 27, wherein the processing circuitry operates to determine an amount of foam in the container, to determine a foam density in the container and to determine the liquid level attributable to the foam, based on the amount of foam and the density of foam. 29. The system according to claim 28, wherein the processing circuitry functions to identify a foam / liquid interface and to determine the amount of foam in the container based on the foam / liquid interface and a geometry of the container. The system according to claim 29, wherein the processing circuitry functions to identify an air / foam interface and to determine the amount of foam in the container, based on the liquid / foam interface, the air / foam interface and the geometry of the container. The system according to claim 28, wherein the processing circuitry operates to determine a wall thickness of the container and to determine the amount of foam in the container, based on the thickness of container walls. 32. The system according to claim 4, wherein the processing circuitry functions to determine a wall thickness of the container and to take into account the thickness of container walls when determining the actual filling level of the container. The system according to claim 32, wherein the processing circuitry functions to take into account variations in a container volume due to the wall thickness of the container. 34. The system according to claim 4, wherein the processing circuitry functions to take the movement of the contents of the container in a gutter when determining the actual filling level thereof. 35. The system according to claim 1, wherein the processing circuitry operates to determine a wall thickness of the container. 36. The system according to claim 35, wherein the processing circuitry functions to take into account variations in a volume of the container due to the thickness of the walls thereof. 37. The system according to claim 1, wherein the processing circuitry functions to take into account the movement of the contents of the container when determining one or more characteristics thereof. 38. The system according to claim 6, wherein the processing circuitry functions to determine whether the moving container is properly pressurized when the container is moved in a non-inverted orientation.