Motion detection device and detection method therefor

Abstract

An operation detection device includes a pixel array, a plurality of comparators, and a plurality of logic check circuits. A comparator is coupled to a pair of pixel columns of the pixel array. The pixel array acquires a reference image and a current image. The comparator generates a reference feature location map relative to the reference image and a current feature location map relative to the current image. The plurality of logic check circuits process the reference feature location map and the current feature location map to detect an action.

Description

Technical Field

[0001] This invention relates to an action detection device, and more particularly to an action detection device and a method thereof that use strong features in an image frame as a reference for determining whether an action is detected. Background Technology

[0002] In traditional motion detection, motion is determined by comparing two light signals acquired sequentially at two exposure times. However, the light signal acquired at the first exposure time decays over time. If the interval between the two exposure times is long enough (e.g., the pixel array is large or the frame rate is low in sleep mode), the light signal acquired at the first exposure time may be completely lost.

[0003] Therefore, there is a need for an alternative motion detection device that is not affected by the long interval between the two exposure times.

[0004] The information disclosed in the background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the relevant information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] The present invention provides a motion detection device and a motion detection method thereof, which uses strong features in a reference image frame (ignoring other features) to generate a feature position map, and determines motion based on the positional changes of the strong features between the feature position maps of the reference image and the current image.

[0006] This invention provides a motion detection device comprising a first pixel, a second pixel, and a comparator. The second pixel is adjacent to the first pixel. The comparator is coupled to the first pixel and the second pixel and is used to output a first comparison result to represent a first signal difference between the first pixel and the second pixel as a reference difference, and to output a second comparison result to represent a second signal difference between the first pixel and the second pixel as a current difference. The first comparison result is obtained by exposing the first pixel and the second pixel with different shutter speeds, and the second comparison result is obtained by exposing the first pixel and the second pixel with the same shutter speed.

[0007] The present invention also provides an action detection device comprising a first pixel, a second pixel, a third pixel, and a comparator. The second pixel and the third pixel are adjacent to each other on both sides of the first pixel. The comparator is not coupled to the first pixel via any switch, is coupled to the second pixel via a second switch, and is coupled to the third pixel via a third switch. The comparator is used to output a first comparison result, indicating the signal difference value between the first pixel and the second pixel as a first reference difference value; output a second comparison result, indicating the signal difference value between the first pixel and the third pixel as a second reference difference value; output a third comparison result, indicating the signal difference value between the first pixel and the second pixel as a first current difference value; and output a fourth comparison result, indicating the signal difference value between the first pixel and the third pixel as a second current difference value. The first comparison result is obtained by exposing the first pixel and the second pixel with different shutter speeds; the second comparison result is obtained by exposing the first pixel and the third pixel with different shutter speeds; the third comparison result is obtained by exposing the first pixel and the second pixel with the same shutter speed; and the fourth comparison result is obtained by exposing the first pixel and the third pixel with the same shutter speed.

[0008] The present invention also provides a method for operating a motion detection device. The motion detection device includes a pixel array, a plurality of comparators respectively coupled to a pair of pixel columns of the pixel array, a plurality of flip-flops each coupled to a comparator, and a plurality of logic checking circuits each coupled to a flip-flop and a corresponding comparator coupled to the flip-flop. The operating method includes the following steps: generating a reference image in which one pixel column of the pair of pixel columns is exposed at a first shutter speed and the other pixel column of the pair of pixel columns is exposed at a second shutter speed different from the first shutter speed; comparing the light signal of the one pixel column of the pair of pixel columns in the reference image with the light signal of the other pixel column of the pair of pixel columns in the reference image using the plurality of comparators to generate a reference feature position map; generating a current image in which one pixel column of the pair of pixel columns is exposed at the first shutter speed and the other pixel column of the pair of pixel columns is exposed at the first shutter speed; comparing the light signal of the one pixel column of the pair of pixel columns in the current image with the light signal of the other pixel column of the pair of pixel columns in the current image using the plurality of comparators to generate a current feature position map; and processing the reference feature position map and the current feature position map using the plurality of flip-flops and the plurality of logic checking circuits to detect action.

[0009] The motion detection device and motion detection method of this invention are suitable for low frame rate applications.

[0010] The motion detection method of this invention can be implemented in the analog phase without going through an analog-to-digital converter (ADC) and digital processing.

[0011] To make the above and other objects, features and advantages of the present invention more apparent, a detailed description will be provided below with reference to the accompanying drawings. Furthermore, in the description of the present invention, the same components are denoted by the same reference numerals, which will be stated herein as well. Attached Figure Description

[0012] Figure 1A This is a schematic diagram of reference images acquired by the motion detection device of the present invention using different shutter speeds in odd-numbered and even-numbered pixel columns of a pixel array; Figure 1B It is based on Figure 1A A schematic diagram of the obtained feature location map; Figure 2 This is a circuit diagram of two adjacent pixels of the pixel array of the motion detection device according to the first embodiment of the present invention; Figure 3 This is a block diagram of the pixel array of the motion detection device according to the first embodiment of the present invention; Figure 4 This is a block diagram of the motion detection device according to the first embodiment of the present invention; Figure 5 This is a schematic diagram of the logic checking circuit of the motion detection device according to the first embodiment of the present invention; Figure 6A It is used with Figure 1A A schematic diagram comparing a reference image with a current image that does not have motion. Figure 6B It is based on Figure 6A A schematic diagram of the obtained feature location map; Figure 6C It is to handle Figure 1B and Figure 6B A schematic diagram of the logical check results of the feature location map; Figure 7A It is used with Figure 1A A schematic diagram comparing the reference image with the current image having action; Figure 7B It is based on Figure 7A A schematic diagram of the obtained feature location map; Figure 7C It is to handle Figure 1B and Figure 7B A schematic diagram of the logical check results of the feature location map; Figure 8AThis is a schematic diagram of a reference image used to calculate a reference feature position map between an even-numbered pixel and two odd-numbered pixels according to the second embodiment of the present invention. Figure 8B It is relative Figure 8A A schematic diagram of the current image, which does not have movement and which has movement from left to right; Figure 8C It is relative Figure 8A A schematic diagram of the current image, which does not have movement and which has movement from right to left; Figure 9 This is a block diagram of the pixel array of the motion detection device according to the second embodiment of the present invention; and Figure 10 This is a block diagram of the motion detection device according to the second embodiment of the present invention.

[0013] Explanation of reference numerals in the attached figures 400, 1000 motion detection devices 31, 91 pixel array 32, 92 comparators 35, 95 processor Detailed Implementation One object of the present invention is to provide an action detection apparatus capable of sequentially generating a reference image and a current image. The reference image is used to generate a feature location map of strong features. The current image is used to check whether these strong features show positional changes to determine whether action is detected.

[0014] In this invention, the reference image can be updated at predetermined time intervals, such as every two frame cycles or every predetermined number of frame cycles. That is, the current image is the next image acquired immediately after the reference image, or an image acquired two frame cycles after the reference image. In this invention, when motion is detected, the motion detection device can be woken up from a sleep mode or a low-power mode and enter a normal mode, but is not limited to these. Alternatively, the reference image is not updated as long as no motion is detected.

[0015] Please refer to Figure 1A and Figure 1B As shown, Figure 1A A schematic diagram of reference images acquired by the motion detection device of the present invention using different shutter speeds in odd-numbered and even-numbered pixel columns of a pixel array; Figure 1B It is based on Figure 1A A schematic diagram of the obtained feature location map. The motion detection device is, for example, an optical mouse, but is not limited thereto.

[0016] In this invention, the motion detection device is used to generate a reference image to compare with a current image to detect motion, the current image being an image frame acquired at a different exposure period than the reference image. Figure 1A In the displayed reference image, the odd-numbered pixel column is exposed at a first shutter speed (e.g., shown as shutter speed 1), and the even-numbered pixel column is exposed at a second shutter speed (e.g., shown as shutter speed 2), which is 10% to 20% longer than the first shutter speed. That is, the shutter speeds of the odd-numbered and even-numbered pixel columns are different. The numerical values ​​in the reference image represent the light signal detected by each pixel. The larger the value, the stronger the detected light.

[0017] Figure 1B The displayed reference feature location map is obtained by comparing odd-numbered pixels with even-numbered pixels using a voltage comparator built into the pixel array (illustrated below). For example, in Figure 1A In the reference image's pixel pair 100A1 (containing one odd pixel and one even pixel), if the odd pixel has a higher light signal than the even pixel, it indicates that the position of the pixel pair 100A1 has strong features. Therefore, as... Figure 1B The reference feature location map 100B1 is assigned the value "1". Conversely, when the odd-numbered pixels have a lower light signal than the even-numbered pixels (e.g., pixel pair 100A2 in the reference image), it indicates that the location of pixel pair 100A2 does not have strong features, therefore... Figure 1B The reference feature location map 100B2 is given the value "0". Other values ​​of the reference feature location map can be obtained using the same method. The horizontal dimension of the reference feature location map is half that of the reference image, for example, changing from 10×10 to 10×5.

[0018] Please refer to Figures 2 to 4 , Figure 2 This is the motion detection device 400 of the first embodiment of the present invention (shown on...). Figure 4 The circuit diagram of two adjacent pixels (shown as odd pixels and odd pixels) of the pixel array 31; Figure 3 This is a block diagram of the pixel array 31 of the motion detection device 400 according to the first embodiment of the present invention; and Figure 4 This is a block diagram of the motion detection device 400 according to the first embodiment of the present invention.

[0019] The motion detection device 400 includes a pixel array 31 and multiple comparators 32. Figure 4 Comparators 1 to 5 are shown as comparators 1 to 5, respectively, coupled to a pair of pixel columns (shown as odd pixel columns and odd pixel columns) of pixel array 31, and each group (e.g. Figure 4(Displayed as FFr1 and FFr2) Multiple flip-flops coupled to a comparator 32, each coupled to a flip-flop, and multiple logic check circuits (e.g., FFr1 and FFr2) coupled to a corresponding comparator 32. Figure 4 The display shows LC1 and LC2), and processor 35. It must be noted that, although... Figures 3 to 4 The display pixel array 31 includes two pixel rows 31r1 and 31r2, two corresponding flip-flops FFr1 and FFr2, and two logic check circuits LC1 and LC2, which are for illustration only and not for limiting the invention. To generate Figure 1A The reference image shows that pixel array 31 comprises ten pixel rows and ten pixel columns, with each pixel row having a corresponding flip-flop row and a corresponding logic check circuit row. That is, every two pixel columns share a comparator, a set of flip-flops, and a set of logic check circuits. The number of flip-flops in each set is equal to the number of pixel rows in pixel array 31, and the number of logic check circuits in each set is equal to the number of pixel rows in pixel array 31.

[0020] The odd pixel (e.g.) Figure 3-4 Displayed as 1st) and the idol elements (e.g.) Figure 3-4 (Displayed as 2nd) represents two adjacent pixels. Each comparator 32 is coupled to an odd pixel and an odd pixel. For example... Figure 4 In the image, comparator 1 is coupled to the first pixel 1st and the second pixel 2nd of pixel row 31r2. The first pixel 1st and the second pixel 2nd are dual-junction transistor (BJT) pixels and each contains a source follower transistor (e.g., Figure 2 Displayed as sf1, sf2) and read transistors (e.g.) Figure 2 (Displayed as RD1, RD2). The drains of the source follower transistors sf1 and sf2 of the first pixel 1st and the second pixel 2nd are coupled to the differential pair of comparator 32. The sources of the read transistors RD1 and RD2 of the first pixel 1st and the second pixel 2nd are coupled to the current source of comparator 32.

[0021] The first embodiment will be described using comparator 1 as an example, and the operation of other comparators is the same, so it will not be described in detail. Comparator 1 outputs the first comparison result (e.g., Figure 1B Displayed as 100B1, 100B2) to represent the first pixel and the second pixel (e.g., ...). Figure 1AThe first signal difference between pixel pairs 100A1 and 100A2 is displayed as a reference difference. When the first comparison results 100B1 and 100B2 are obtained, the first pixel (e.g., the left pixel of 100A1 and 100A2) is exposed at a first shutter speed, and the second pixel (e.g., the right pixel of 100A1 and 100A2) is exposed at a second shutter speed different from the first shutter speed. As described above, the second shutter speed is 10% to 20% longer than the first shutter speed to distinguish whether the difference is strong enough. The flip-flop FFr2 is used to record the first comparison results 100B1 and 100B2. When comparator 1 outputs a new first comparison result for the first pixel and the second pixel of pixel row 31r1, the previous first comparison result in comparator FFr2 is moved to comparator FFr1 and the new first comparison result overwrites comparator FFr2, that is, the plurality of flip-flops operate in a first-in-first-out (FIFO) order.

[0022] Please refer to Figure 6A and Figure 7A , Figure 6A It is used with Figure 1A A schematic diagram comparing the reference image with the current image without action, obtained from pixel array 31; Figure 7A It is used with Figure 1A A schematic diagram comparing the reference image with the current image with motion obtained from pixel array 31. Figure 6B and Figure 7B It is based on respectively Figure 6A and Figure 7A The obtained feature location map.

[0023] After the pixel array 31 acquires the current image, comparator 1 then outputs a second comparison result (e.g., Figure 6B It is displayed as 600B. Figure 7B Displayed as 700B) to represent the first pixel and the second pixel (e.g., ...). Figure 6A Displayed as pixel pair 600A, Figure 7A The second signal difference between pixel pairs 700A and 700B is displayed as the current difference. Upon obtaining the second comparison results 600B and 700B, the first pixel (e.g., the left pixel of 600A and 700A) is exposed at the first shutter speed, and the second pixel (e.g., the right pixel of 600A and 700A) is also exposed at the first shutter speed. That is, all pixels of pixel array 31 are exposed at the same shutter speed to obtain the current image.

[0024] The logic check circuit LC2 is coupled to comparator 1 and flip-flop FFr2. (Refer to...) Figure 5Each logic checking circuit includes an inverter 51 and an AND gate 52. For example, the input of the inverter 51 of the logic checking circuit LC2 is used to receive the second comparison result (e.g., displayed as CPM_31r2). The first input of the AND gate 52 of the logic checking circuit LC2 is used to receive the first comparison result (e.g., displayed as RPM_31r2) from the flip-flop FF2, and the second input of the AND gate 52 of the logic checking circuit LC2 is coupled to the output of the inverter 51 of the logic checking circuit LC2. In this invention, the first comparison result RPM_31r2 and the second comparison result CPM_31r2 are obtained relative to different image frames, that is, the reference image and the current image captured by the pixel array 31. More specifically, the first comparison result RPM_31r2 is only transmitted to the comparator FFr2 but not to the logic checking circuit LC2, and the second comparison result CPM_31r2 is only transmitted to the logic checking circuit LC2 but not to the comparator FFr2, for example, by controlling a switch.

[0025] The operation of other pixel columns, other comparators (e.g., comparators 2 to 5), other flip-flops, and other logic check circuits is the same as that of the first and second pixel columns, comparator 1, flip-flop FFr2, and logic check circuit LC2, except for the time difference (e.g., controlled by the row selection signal and read signal generated by the timing controller). After understanding the above description, you can also understand their operation.

[0026] Please refer to again Figure 4 The pixel array 31 comprises multiple pairs of first and second pixels arranged in a matrix. Each of the plurality of comparators 32 is coupled to a column of first and second pixel pairs in the matrix. Thereby, the plurality of comparators (e.g., comparators 1 to 5) outputs a plurality of first comparison results as... Figure 1B The displayed reference feature location map and the output of multiple second comparison results serve as... Figure 6B Figure 7B The current feature location map is displayed. The multiple logic check circuits process... Figure 1B Reference feature location map and Figure 6B and Figure 7B The current feature location map is used to determine the action. For example, Figure 6C Processed by the aforementioned multiple logic check circuits Figure 1B and Figure 6B A schematic diagram of the logical check results of the feature location map; Figure 7C Processed by the aforementioned multiple logic check circuits Figure 1B and Figure 7B A schematic diagram of the logical check results of the feature location map.

[0027] Processor 35, such as a microcontroller unit (MCU), application-specific integrated circuit (ASIC), or programmable logic array (FPGA), is used to control the opening and closing of the switching device and to determine actions. Processor 35 receives... Figure 1B The processor 35 obtains a reference feature location map, identifies and records the positions of strong features in the reference feature location map (e.g., marked as "1"). The processor 35 then determines the location of strong features based on a first comparison result showing strong features in the reference feature location map. Figure 6B and Figure 7B The number of second comparison results processed by the logic checking circuit in the current feature location map. For example, processor 35 controls the logic checking circuit for positions related to "1" in the reference feature location map to be activated, and controls the logic checking circuit for positions related to "0" in the reference feature location map to be deactivated. That is, Figure 6B and Figure 7B In the middle, only the area with the black border is visible. Figure 1B The strong features are processed by the corresponding logic checking circuit, while other locations are ignored.

[0028] like Figure 6C As shown, if no action is detected, the logic check results output by the multiple logic check circuits only contain "0". Figure 7C As shown, if an action is detected, the logic check results output by the plurality of logic check circuits contain some "1". In one embodiment, for noise reduction purposes, the processor 35 confirms an action when the number of "1"s in the logic check results exceeds a predetermined threshold. In another embodiment, strong features can be represented by "0".

[0029] To save power, the multiple comparators 32 are configured only for pixel pairs in the first and second pixel columns relative to a portion (e.g., 1 / 2 or 1 / 3) of the pixel matrix of the pixel array 31. That is, not every pixel column of the pixel array 31 is used to detect action.

[0030] In another embodiment, the processor 35 determines that a portion of the strong features of the reference feature location map are processed by logic checking circuits. For example, the processor 35 includes a counter for counting a predetermined number of strong features, such as ten strong features, and the processor 35 enables ten logic checking circuits associated with the ten strong features to process their first comparison result and second comparison result. This also reduces power consumption. That is, in this invention, the processor 35 can adaptively select based on the content of the reference image. Figure 1B Reference feature location map and Figure 6B and Figure 7B The area in the current feature location map that is processed by the logic check circuit, for example, processor 35 is configured to receive Out1 to Out5 from comparators 1 to 5 for selection.

[0031] To improve detection sensitivity, the comparator of the motion detection device of the present invention also uses switching control technology to compare a column of pixels (e.g., Figures 8A-8C (even-numbered pixel columns) and two adjacent pixel columns (e.g.) Figures 8A-8C (Odd-numbered pixel columns).

[0032] Please refer to Figure 9 This is the motion detection device 1000 of the second embodiment of the present invention (shown in...). Figure 10 A block diagram of a pixel array. The difference between the second embodiment and the first embodiment is that the second embodiment also includes a plurality of first switches SW1 and a plurality of second switches SW2 for connecting or disconnecting odd-numbered pixel columns, for example, shown as second pixel 2nd or third pixel 3rd, and comparator 92.

[0033] The motion detection device 1000 includes a pixel array 91 and multiple comparators 92 (e.g., ...). Figure 10 Displayed as comparators 1 to 5), multiple flip-flops (e.g., ... Figure 10 Displayed as FFr1, FFr2, FFr1′, FFr2′) and multiple logic check circuits (e.g. Figure 10 (Displayed as LC1, LC2, LC1′, LC2′).

[0034] The second embodiment will also be described using the first pixel 1st, the second pixel 2nd, and the third pixel 3rd as examples. Those skilled in the art, after understanding the operation of pixels 1st to 3rd, will also be able to understand the operation of other pixels.

[0035] The second pixel (2nd) and the third pixel (3rd) are adjacent to the first pixel (1st) on either side of it. For example, Figures 9-10 The first pixel (1st) is an even number of pixels, and the second pixel (2nd) and the third pixel (3rd) are odd numbers of pixels, but this is not the only case.

[0036] The second embodiment is also described using comparator 1 as an example, and the operation of other comparators can be understood based on the operation of comparator 1. Comparator 1 is not coupled to the first pixel 1st through any switches, but is coupled to the second pixel 2nd through a second switch and to the third pixel 3rd through a third switch. Comparator 1 outputs a first comparison result (e.g., Figure 8A The position of the reference feature location map shown (801R) indicates the first pixel and the second pixel (e.g., Figure 8A The signal difference between pixel pairs (800AL) is used as a first reference difference. For example, switches SW1 and SW2 connect the second pixel 2nd to comparator 1 and disconnect the third pixel 3rd from comparator 1. Comparator 1 also outputs a second comparison result (e.g., ...). Figure 8A The position of the reference feature location map shown (802R) indicates the first pixel and the third pixel (e.g., Figure 8A The signal difference between pixel pairs (800AR) is used as a second reference difference. For example, switches SW1 and SW2 connect the third pixel 3rd to comparator 1 and disconnect the second pixel 2nd from comparator 1. Similarly, when obtaining the first comparison result 801R, the first pixel 1st is exposed at a first shutter speed while the second pixel 2nd is exposed at a second shutter speed different from the first shutter speed. When obtaining the second comparison result 802R, the first pixel 1st is exposed at the first shutter speed while the third pixel 3rd is exposed at the second shutter speed. In one embodiment, the second shutter speed is 10% to 20% longer than the first shutter speed. Figure 8A Only the first comparison result 801R and the second comparison result 802R of the reference feature location map are displayed. By using... Figure 1A and Figure 1B Using the same method, a reference feature location map can be obtained when comparator 1 is connected to an odd pixel (e.g., the second pixel 2nd) to the left of an even pixel (e.g., the first pixel 1st), and another reference feature location map can be obtained when comparator 1 is connected to an odd pixel (e.g., the third pixel 3rd) to the right of an even pixel (e.g., the first pixel 1st).

[0037] Comparator 1 outputs the third comparison result (e.g.) Figure 8B and Figure 8C The current feature location map shown is at position 801C to indicate the first pixel and the second pixel (e.g., ...). Figure 8B Pixels for 800BL and Figure 8C The signal difference between pixels (800CL) is used as the first current difference. For example, switches SW1 and SW2 connect the second pixel (2nd) to comparator 1 and disconnect the third pixel (3rd) from comparator 1. Comparator 1 also outputs a fourth comparison result (e.g., ...). Figure 8B and 8C The current feature location map shown is at position 802C to represent the first pixel and the three pixels (e.g., ...). Figure 8B Pixels for 800BR and Figure 8C The signal difference between pixel pairs (800CR) is used as the second current difference. For example, switches SW1 and SW2 connect the third pixel 3rd to comparator 1 and disconnect the second pixel 2nd from comparator 1. Similarly, when obtaining the third comparison result 801C and the fourth comparison result 802C, the first pixel 1st, the second pixel 2nd, and the third pixel 3rd are exposed at the same shutter speed, for example, the first shutter speed.

[0038] Figure 8B and Figure 8C Only the third comparison result 801C and the fourth comparison result 802C of the current feature location map are displayed. This can be achieved by using methods such as... Figures 6A-6B and Figures 7A-7B Using the same method, a current feature position map can be obtained when comparator 1 is connected to an odd pixel (e.g., the second pixel 2nd) to the left of an even pixel (e.g., the first pixel 1st), and another current feature position map can be obtained when comparator 1 is connected to an odd pixel (e.g., the third pixel 3rd) to the right of an even pixel (e.g., the first pixel 1st).

[0039] In the second embodiment, the first comparison result 801R and the second comparison result 802R are obtained from the first image frame, and the third comparison result 801C and the fourth comparison result 802C are obtained from a second image frame that is different from the first image frame.

[0040] A first flip-flop (e.g., FFr2) is coupled to comparator 1 via a first output switch to receive a first comparison result 801R. A second flip-flop (e.g., FFr2') is coupled to comparator 1 via a second output switch to receive a second comparison result 802R. A first logic check circuit (e.g., LC2) is coupled to the first flip-flop FFr2 and to comparator 1 via the first output switch. A second logic check circuit (e.g., LC2') is coupled to the second flip-flop FFr2' and to comparator 1 via the second output switch. Preferably, the first output switch and the second output switch are not turned on simultaneously. In one embodiment, when the second pixel 2nd is coupled to comparator 1, the output of comparator 1 is coupled to FFr2 or LC2; and when the third pixel is coupled to comparator 1, the output of comparator 1 is coupled to FFr2' or LC2'. Thus, the plurality of flip-flops can store two reference feature position maps, and the plurality of logic check circuits can receive two current feature position maps.

[0041] In implementations that detect motion using only pixel pairs 800AR and 800BR / 800CR, similar to the first embodiment, although left-to-right movement can be detected, for example... Figure 8A In 802R, the "1" is changed to Figure 8B The value is "0" in 802C, but movement from right to left cannot be detected, for example. Figure 8C In the second embodiment, the pixel 802C remains "1", indicating no action. By examining the other pixel pair 800AL and 800BL / 800CL, both left-to-right and right-to-left movements can be detected, thus improving detection sensitivity.

[0042] In the second embodiment, the processor 95 is also used to control the switching elements and determine and record the strong features in the reference feature location map, so that only the positions with relatively strong features in the current feature location map are processed by multiple flip-flops and multiple logic check circuits, similar to the operation of the processor 35 in the first embodiment.

[0043] It must be noted that, although Figures 9-10 The diagram shows that the even-numbered pixels are not coupled to comparator 92 via a switch, but the invention is not limited thereto. In other embodiments, the odd-numbered pixels are not coupled to comparator 92 via a switch, while the even-numbered pixels are coupled to comparator 92 via a switch. In yet another embodiment, all pixels are coupled to comparator 92 via separate switches, and processor 95 controls the opening and closing of the switches.

[0044] Please refer to again Figure 1A-1B , Figures 6A-6C and Figures 7A-7C The operation method of the motion detection device 400 in this embodiment of the invention includes the following steps: generating a reference image (e.g., Figure 1A In the reference image, one pixel column (e.g., an odd-numbered pixel column) of a pair of pixel columns is exposed at a first shutter speed, while the other pixel column of the pair (e.g., an even-numbered pixel column) is exposed at a second shutter speed different from the first shutter speed; using multiple comparators (e.g. Figure 4 Comparators 1 to 5 compare the light signal of one pixel column of the pair of pixel columns in the reference image with the light signal of the other pixel column of the pair of pixel columns in the reference image to generate a reference feature location map (e.g., Figure 1B ); generate the current image (e.g. Figure 6A and Figure 7A The current image shows one pixel column (e.g., an odd-numbered pixel column) exposed at the first shutter speed and the other pixel column (e.g., an even-numbered pixel column) exposed at the first shutter speed; the plurality of comparators compare the light signal of one pixel column of the current image with the light signal of the other pixel column of the current image to generate a current feature location map (e.g. Figure 6B and Figure 7B ); and with multiple flip-flops (e.g. Figure 4 The displayed FFR1, FFR2) and multiple logic check circuits (e.g., FFR1, FFR2) and multiple logic check circuits Figure 4 The displayed LC1 and LC2) process the reference feature position map and the current feature position map to detect actions. The operating method is also applicable to... Figures 8A to 8C In the second embodiment, two reference feature position maps and two current feature position maps are generated according to the state of the switch.

[0045] In this invention, a pair of pixel columns in the pixel array may include an even-numbered pixel column and an odd-numbered pixel column as described in the first embodiment, or an even-numbered pixel column and two odd-numbered pixel columns as described in the second embodiment, or an odd-numbered pixel column and two even-numbered pixel columns.

[0046] As described above, the processor 35 / 95 marks strong features in the reference feature location map and ignores locations in the reference feature location map (and the current feature location map) that do not have strong features.

[0047] To save power, the multiple logic checking circuits only process (controlled by processor 35 / 95) a portion of the strong features in the reference feature location map and the positions of the current feature location map relative to the portion of strong features. That is, when detecting motion, not all strong features need to be considered.

[0048] It must be noted that, although Figure 2 Each pixel contains a second circuit group (e.g., displayed as 29). odd and 29 even The storage of another optical signal to the storage capacitor INTD (e.g., with a different exposure time than the optical signal stored in the storage capacitor INT) is for illustrative purposes only and not for limiting the invention. In other embodiments, the pixels of pixel array 31 / 91 do not include the second circuit group 29. odd and 29 even In another embodiment, the upper half of the comparator is coupled to the drain of the source follower couplers sf3 and sf4, and the lower half of the comparator is coupled to OUTD.

[0049] It must be noted that although the strong feature is described in this invention description as odd / even pixels being brighter than even / odd pixels, this is for illustrative purposes only and not for limiting the invention. In other embodiments, the strong feature is defined as odd / even pixels being darker than even / odd pixels, which can be achieved by changing the connection between the drain of the source follower sf3 and sf4 and the comparator.

[0050] In traditional motion detectors, the sensor requires three stages of processing to confirm a frame. These three stages include shutter exposure, data transfer, and signal processing. However, in this invention, there is no need to transfer pixel data from analog circuitry to digital circuitry, nor is post-processing of pixel data required. Therefore, the energy consumed in processing pixel data can be saved. Furthermore, in this invention, the comparator / amplifier is built into the pixel.

[0051] It must be noted that although the description in this invention describes comparing the position map and calculating the logic check results during the analog phase, this is for illustrative purposes only and not for limiting the invention. In other embodiments, the position map may first be converted (e.g., using an ADC) to the digital phase, and then the processor obtains the logic check results in the digital phase.

[0052] In summary, because conventional motion detectors store the earlier acquired light signal into the pixel capacitor for comparison with the later acquired light signal, the energy of the earlier stored light signal decays between the two exposure times. Therefore, this invention proposes a different motion detection device (e.g., see reference 1). Figure 4 and Figure 10 ) and its motion detection methods (e.g., refer to Figure 1A-1B , Figures 7A-7C and Figures 8A-8C The system stores the location map of strong features in multiple flip-flops for comparison with the new location map during the simulation phase to obtain the logical check results of the position changes of the strong features. If no action occurs, these strong features will be located in the same position in the new location map; when an action occurs, these strong features will appear in a different position in the new location map. Since the multiple flip-flops only record "1" or "0", the problem of energy decay can be eliminated.

[0053] While the present invention has been disclosed through the foregoing examples, it is not intended to limit the invention. Anyone skilled in the art to which this invention pertains can make various modifications and alterations without departing from the spirit and scope of the invention. Therefore, the scope of protection of this invention shall be determined by the appended claims.

Claims

1. A motion detection device, the motion detection device comprising: First pixel; The second pixel is adjacent to the first pixel; as well as A comparator, coupled to the first pixel and the second pixel, and used for... Output a first comparison result to represent the first signal difference between the first pixel and the second pixel as a reference difference. Output a second comparison result to represent the second signal difference between the first pixel and the second pixel as the current difference, wherein, The first comparison result is obtained by exposing the first pixel and the second pixel with different shutter speeds, and The second comparison result is obtained by exposing the first pixel and the second pixel with the same shutter speed.

2. The motion detection device according to claim 1, wherein, The first pixel and the second pixel respectively include: A source follower transistor, wherein the drain of the source follower transistor of the first pixel and the second pixel is coupled to the differential pair of the comparator; and The read transistors of the first pixel and the second pixel are coupled to the current source of the comparator.

3. The motion detection device according to claim 1, wherein, The first comparison result and the second comparison result were obtained from different image frames.

4. The motion detection device according to claim 1, wherein, When acquiring the first comparison result, the first pixel is exposed at a first shutter speed and the second pixel is exposed at a second shutter speed, wherein the second shutter speed is 10% to 20% longer than the first shutter speed. When obtaining the second comparison result, the same shutter speed is equal to the first shutter speed.

5. The motion detection device according to claim 1, further comprising: A flip-flop is coupled to the comparator and used to record the first comparison result.

6. The motion detection device according to claim 5, further comprising: A logic checking circuit is coupled to the comparator and the flip-flop.

7. The motion detection device according to claim 6, wherein, The logic checking circuit includes: An inverter, the input of which is used to receive the second comparison result, and An AND gate, wherein the first input of the AND gate is used to receive the first comparison result from the inverter, and the second input of the AND gate is coupled to the output of the inverter.

8. The motion detection device according to claim 1, wherein the motion detection device comprises: Multiple pairs of the first pixel and the second pixel are arranged in a matrix; and Multiple comparators, each of which is used to couple multiple pairs of the first pixel and a column of the second pixel in the matrix. in, The multiple comparators are used to output multiple first comparison results as a reference feature location map and to output multiple second comparison results as the current feature location map.

9. The motion detection device according to claim 8, wherein, The plurality of comparators are configured only relative to a subset of columns of the first and second pixels in the matrix.

10. The motion detection device according to claim 8, further comprising: Multiple logic checking circuits are used to process the reference feature location map and the current feature location map; and A processor configured to determine, based on a first comparison result representing a strong feature in the reference feature location map, the number of second comparison results processed by the plurality of logic checking circuits in the current feature location map.

11. A motion detection device, the motion detection device comprising: First pixel; The second pixel and the third pixel are adjacent to each other on both sides of the first pixel; A comparator, which is not coupled to the first pixel via any switch, is coupled to the second pixel via a second switch, and is coupled to the third pixel via a third switch, is used for... Output a first comparison result, indicating that the signal difference value between the first pixel and the second pixel is used as a first reference difference value. Output a second comparison result, indicating that the signal difference value between the first pixel and the third pixel is used as a second reference difference value. The third comparison result is output, indicating that the signal difference value between the first pixel and the second pixel is used as the first current difference value. The fourth comparison result is output, indicating that the signal difference value between the first pixel and the third pixel is used as the second current difference value, wherein... The first comparison result is obtained by exposing the first pixel and the second pixel with different shutter speeds. The second comparison result is obtained by exposing the first pixel and the third pixel with different shutter speeds. The third comparison result is obtained by exposing the first pixel and the second pixel with the same shutter speed, and The fourth comparison result is obtained by exposing the first pixel and the third pixel with the same shutter speed.

12. The motion detection device according to claim 11, wherein, The first comparison result and the second comparison result are obtained from the first image frame, and The third comparison result and the fourth comparison result are obtained from a second image frame that is different from the first image frame.

13. The motion detection device according to claim 11, wherein, When acquiring the first comparison result, the first pixel is exposed at a first shutter speed and the second pixel is exposed at a second shutter speed. When obtaining the second comparison result, the first pixel is exposed at the first shutter speed and the third pixel is exposed at the second shutter speed, and The second shutter speed is 10% to 20% longer than the first shutter speed.

14. The motion detection device according to claim 11, further comprising: A first flip-flop, the first flip-flop being coupled to the comparator via a first output switch to receive the first comparison result; and A second flip-flop is coupled to the comparator via a second output switch to receive the second comparison result.

15. The motion detection device according to claim 14, further comprising: A first logic checking circuit, which is connected to the first flip-flop and coupled to the comparator via the first output switch; and A second logic check circuit is connected to the second flip-flop and coupled to the comparator via the second output switch.

16. A method of operating a motion detection device, the motion detection device comprising a pixel array, a plurality of comparators respectively coupled to a pair of pixel columns of the pixel array, a plurality of flip-flops each coupled to a comparator, and a plurality of logic checking circuits each coupled to a flip-flop and a corresponding comparator coupled to the flip-flop, the method comprising: A reference image is generated in which one pixel column of the pair of pixel columns is exposed at a first shutter speed and the other pixel column of the pair of pixel columns is exposed at a second shutter speed different from the first shutter speed; The optical signal of one pixel column of the pair of pixel columns in the reference image is compared with the optical signal of the other pixel column of the pair of pixel columns in the reference image using the plurality of comparators to generate a reference feature location map; Generate a current image in which one pixel column of the pair of pixel columns is exposed at the first shutter speed and the other pixel column of the pair of pixel columns is exposed at the first shutter speed; The optical signal of one pixel column of the pair of pixel columns in the current image is compared with the optical signal of the other pixel column of the pair of pixel columns in the current image using the plurality of comparators to generate a current feature location map; as well as The reference feature position map and the current feature position map are processed by the plurality of flip-flops and the plurality of logic checking circuits to detect the action.

17. The operating method according to claim 16, wherein, The second shutter speed is 10% to 20% longer than the first shutter speed.

18. The operating method according to claim 16, further comprising: Mark the strong features in the reference feature location map; and Locations in the reference feature location map that do not have the strong feature are ignored.

19. The operating method according to claim 18, wherein, When detecting the action, the plurality of logic checking circuits only process a portion of the strong features in the reference feature location map and the position of the current feature location map relative to the portion of strong features.

20. The operating method according to claim 16, wherein, The pair of pixel columns in the pixel array includes: A column of even pixels and a column of odd pixels, or One column of even-numbered pixels and two columns of odd-numbered pixels.

Images