JP2006277315A - Pattern recognition device, pattern recognition method, and electronic equipment provided with the pattern recognition device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily execute recognition in a pattern recognition device including bar code and character readers while improving the recognition rate thereof. <P>SOLUTION: This device comprises an image analysis means (CPU 19) analyzing an image inputted from an image input part (camera module 10); a correction means (CPU 19) determining a quality adjustment parameter necessary for improvement in recognition rate based on the analysis result of the image by the image analysis means, and feeding back the image quality adjustment parameter to the image input part 10 to correct the image; and a recognition means (CPU 19) performing, after correcting the image by the correction means, pattern recognition of the corrected image. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a pattern recognition device such as a barcode reader, a pattern recognition method, and an electronic apparatus provided with the pattern recognition device, and in particular, a barcode reader, a barcode reading method, and a barcode reader thereof with improved recognition rate. The present invention relates to an electronic device.

  In recent years, mobile phones are equipped with small camera modules, and two-dimensional barcodes such as QR codes are photographed and data (URLs, images, sounds, texts) stored in the barcodes are read or represented by OCR. In this way, products that can read character information directly have been developed and put on the market. There are various types of barcodes and character information, such as their methods, sizes, and pitches, and there is a need to improve the recognition rate and shorten the reading time. In order to improve the recognition rate and shorten the recognition time, it is necessary to improve the basic performance of the camera (resolution, contrast, etc.), improve the performance of the reading engine, and improve the reading method and sequence. The image quality of the captured image greatly affects the recognition rate.

  Usually, the camera is equipped with brightness adjustment, edge enhancement, and contrast adjustment functions by AE (Automatic Exposure), and adjustments are made to improve the recognition rate. Reading performance is also recognized by the recognition engine according to the image quality of the camera. The engine tuning for improvement is performed.

  However, considering the scene where the barcode is read by the camera, there are a wide variety of external factors that affect the image quality, such as lighting conditions, barcode printing status, shooting distance, and focus. It is difficult to realize the adjustment to get the recognition, and even the same code does not necessarily recognize well. For example, Patent Document 1 (Conventional Example 1) proposes an apparatus intended to improve the recognition rate of barcodes, characters, and the like.

  FIG. 7 is a block diagram showing the configuration of the image reading apparatus of Conventional Example 1, and FIG. The image reading apparatus of Conventional Example 1 includes a control unit 31, a memory 20a, a state detection unit 32, an operation control unit 33, an image input unit 34, an image processing unit 35, an image compression unit 36, an external An I / F (interface) 15a, an operation unit 37, and a bus 18a are provided.

  The control unit 31 controls the entire apparatus, and the memory 20a stores a control program and various data necessary for the control. The state detection unit 32 inspects the operation state and operation state of the apparatus, and the operation control unit 33 controls device operations such as a document reading operation. The image input unit 34 scans the image and converts it into an electrical signal, and the image processing unit 35 performs the read image digital processing. The image compression unit 36 performs image data compression processing, and the external I / F 15a interfaces with a computer for setting reading conditions and transferring read data. The operation unit 37 inputs an operation by the operator and transmits information to the operator, and the bus 18a connects these units.

  In the operation flowchart of FIG. 8, first, in step S21, the scanner distribution function is selected. Next, a document to be read is set on an ADF (automatic document feeder) or contact glass (S22), an automatic reading condition mode provided as a menu on the screen of the apparatus operation unit is selected, and necessary setting items are selected. Start (S23). Then, the first reading condition among the preset conditions N is set and the pre-scan is executed (S24). The read image is temporarily stored in the memory (S25). The stored image is subjected to OCR processing using the OCR engine of the image processing library, and the number of recognized characters is counted and described (S26). These prescan and OCR processes are repeated up to the Nth and the results are recorded (S27). Here, the condition with the largest number of characters recognized by OCR is selected (S28). A main scan is performed under the selected reading condition, and the image input unit performs reading (S29). The main scanned image is distributed to the designated distribution destination (S30).

  The reading conditions in this case are MTF correction intensity, smoothing correction intensity, resolution, contrast, luminance, binarization threshold value, and the like.

  This device can set the optimum reading condition by a simple method by selecting the reading condition of the scanner, pre-scanning, and OCR processing the image to automatically find the reading condition with a high recognition rate. It is an image reading device that can.

  Further, an image diagnosis support apparatus (CAD: Computer Aided Diagnosis) controlled by a plurality of reading conditions is disclosed in Patent Document (Conventional Example 2).

  FIG. 9 is a block diagram showing the configuration of the conventional image diagnosis support apparatus of FIG. 2, and FIG. 10 is a flowchart for explaining the operation of this apparatus. The image diagnosis support apparatus includes an input unit 41, an image input unit 34a, a processing unit 42, a storage unit 20b, a display unit 21b, and a temporary storage unit 20c. The processing unit 42 includes an image data characteristic detection unit 51, a threshold value correction unit 52, and a diagnosis unit 53. The storage unit 20b includes a basic program 55, an image data characteristic detection program 56, a threshold value 57, and a threshold value. A value correction table 58 is stored.

  The input unit 41 is a part for inputting data using a keyboard or the like, and the input data is supplied to the processing unit 42. The image input unit 34 a is a part for inputting medical image data such as an X-ray film, and the image data is also supplied to the processing unit 42.

  The processing unit 42 is configured by a CPU or the like and executes various processes. Specifically, the processing unit 42 executes the basic program 55 and various application programs according to instructions from the input unit 41, and stores the results in the temporary storage unit 20c or displays them on the display unit 21b.

    The image data characteristic detection unit 51 detects the image characteristic of the image data according to the image data characteristic detection program 56. Based on the threshold correction table 58, the threshold correction unit 52 determines and corrects a threshold correction value corresponding to the detected image characteristic. The diagnosis unit 53 uses the corrected threshold value to detect an abnormal part of the image data and diagnose a medical condition.

  The image data characteristic detection program 56 is a subprogram for detecting the image characteristic of the image data, and determines the characteristic level from the detected spatial frequency characteristic. The image characteristics may be gradation characteristics, resolution, brightness, and density. The threshold value 57 is used when the diagnosis unit 53 performs image data pattern recognition. The threshold value correction table 58 stores the characteristic level of the image data detected by the image data characteristic detection unit 51 and the correction value of the threshold value in association with each other.

  In the flowchart of FIG. 10, first, image data is input from the image input unit 34a and output to the processing unit 42 (step S41). Then, the image data characteristic detection unit 51 executes processing according to the image data characteristic detection program 56 to detect the image characteristic (step S42). The image data characteristic detection unit 51 performs Fourier transform on the input image data to perform spatial frequency conversion, and analyzes the magnitude of the spatial frequency component. The characteristic level is determined based on the magnitude of the spatial frequency component.

  Next, the threshold value correction unit 52 determines a correction value corresponding to the characteristic level determined in the image data characteristic detection process based on the threshold value correction table 58 (step S43). The threshold value correcting unit 52 calculates a new threshold value using the threshold value stored in the threshold value 57 and the correction value (step S44). Then, the diagnosis unit 53 recognizes the pattern of the image data according to the threshold value in step S44, detects an abnormal part of the image data, and diagnoses it (step S45).

  The processing unit 42 determines whether or not abnormality diagnosis instruction data is input from the input unit 41 (step S46). When the instruction data is input (yes), the processing unit 42 displays the threshold value input screen on the display unit 21b. When the instruction data is input (step S47), the processing unit 42 stores the input threshold value in the threshold value 57, and returns to step S44. On the other hand, when the instruction data is not input (no), the processing unit 42 ends the process.

  The diagnostic apparatus corrects the threshold value based on the characteristics of the image data, and detects an abnormal portion of the image data using the corrected threshold value, thereby causing variations in the image data characteristics of the plurality of image data. However, it is possible to detect an abnormal part under uniform conditions.

JP 2004-234261 A JP 2004-113281 A

  However, in the configuration of Conventional Example 1 described above, one of a plurality of reading conditions such as resolution and contrast is selected to find a reading condition with a high recognition rate, and the recognition rate is automatically determined from the plurality of reading conditions. Since it has not improved, there is a problem that some processing takes time. In addition, there is a problem in that it takes time and effort to perform a pre-scan for finding a reading condition with a high recognition rate, and a memory for storing data must be secured. In particular, when it is installed in a cellular phone where usability and operability are important, pre-scanning time and effort greatly reduce the merchantability. Further, in order to perform pre-scanning, it is necessary to set some reading conditions in advance. However, depending on how to determine the setting value, it cannot be said that an optimal solution is necessarily found by the recognition rate improvement work by pre-scanning.

  In the configuration of Conventional Example 2 described above, the image diagnosis support apparatus detects the characteristics of the image data, corrects the threshold value, and attempts to improve the recognition rate of the corrected image diagnosis. There is a description to correct the parameters. However, these correction parameters are not automatically determined and do not calculate a correction value for each parameter, so that there is a problem that processing time is required.

  An object of the present invention is to provide an image quality adjustment parameter determined to be necessary for improving the recognition rate based on the analysis result of the recognition engine of the target image taken by the camera to solve the above-described problems and feed back to the camera side. In addition, by automatically performing image quality adjustment that is optimal for reading barcodes and the like, pattern recognition that can obtain the optimum recognition rate even in various shooting environments without causing an increase in memory capacity An apparatus, a pattern recognition method, and an electronic apparatus having the pattern recognition apparatus.

  The configuration of the pattern recognition apparatus of the present invention includes an image signal generating means for generating an image signal by photographing, an image for analyzing the image signal from the image signal generating means, and determining and outputting an image quality adjustment parameter based on the analysis result. Analysis means, correction means for correcting the image signal from the image signal generation means based on image quality adjustment parameters from the image analysis means, recognition means for recognizing the corrected image corrected by the correction means, It is characterized by providing.

  Another pattern recognition apparatus according to the present invention includes an image pickup unit and an image that converts an analog signal output from the image pickup unit into a digital signal and performs predetermined image processing on the digital signal to generate a digital video signal. A camera unit that includes a processing unit and a control unit that controls the image processing unit; and a data processing unit that is connected to the camera unit and that executes an image pattern recognition process based on the digital video signal. The processing unit generates an image quality adjustment parameter based on an analysis result for recognizing the digital video signal and feeds back to the camera unit. Further, the camera unit automatically performs image quality adjustment based on the image quality adjustment parameter. It is characterized by correcting.

  According to still another pattern recognition apparatus of the present invention, an image signal generating unit that generates an image signal by photographing, an image signal from the image signal generating unit are analyzed, and an image quality adjustment parameter is determined based on the analysis result. An image analysis means for outputting, an image processing means for correcting the image quality of the image signal so as to improve the image quality by a predetermined image processing program based on an image quality adjustment parameter from the image analysis means, and a corrected image corrected by the image processing means And a pattern recognition means.

  In the present invention, the image quality adjustment parameter can be an image resolution or a contrast, and the image resolution is obtained by performing a second derivative operation on the input image signal to extract a contour signal, and the level of the contour signal is a video. It can be controlled so as to increase with respect to the signal level, and the contrast of the image can be controlled such that the histogram distribution is analyzed with respect to the signal level of the input image signal and the histogram distribution becomes a steep pattern. Also, the pattern recognition device can be a bar code reader.

  The configuration of the electronic device according to the present invention is characterized by including the pattern recognition device described above, and the electronic device can be a portable terminal with a camera capable of reading a barcode.

  The configuration of the pattern recognition method of the present invention is such that the image signal from the image signal generation means is analyzed by the image analysis means, the image quality adjustment parameter is determined and output based on the analysis result, and the image quality adjustment from this image analysis means is performed. The image signal from the image signal generating unit is corrected by a correcting unit based on a parameter, and the corrected corrected image is recognized by a recognition unit.

  The configuration of another pattern recognition method of the present invention is to analyze the image signal from the image signal generation means by the image analysis means, determine and output an image quality adjustment parameter based on the analysis result, and by the image processing means, The image signal is subjected to image quality correction by a predetermined image processing program using an image quality adjustment parameter from the image analysis unit, and the corrected image corrected by the image processing unit is subjected to pattern recognition by a recognition unit. .

  As described above, according to the present invention, since the image quality adjustment optimum for reading the image quality adjustment parameter of the camera is automatically performed based on the image analysis result, the pattern recognition including the barcode, the character reader, etc. There is an effect that the recognition rate of the device can be easily improved.

  Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of an embodiment of the present invention. This embodiment shows a barcode reader mounted on a camera-equipped mobile phone as one of pattern recognition devices. In FIG. 1, a target image (barcode) is taken by a camera (camera module 10). The target image is analyzed by the recognition engine for the image quality adjustment parameter determined to be necessary for improving the image recognition rate. Based on the analysis result, an image quality adjustment parameter determined to be necessary is fed back to the camera side, and image quality adjustment optimal for reading is automatically performed. Therefore, according to this configuration, the recognition rate can be improved.

  1, a camera module 10 serving as a camera unit includes an optical lens 11, an image sensor 12, an AD converter 13, an image processing DSP (Digital Signal Processor) 14, an I / F circuit (interface) 15, a power driver circuit 16, It consists of a microcomputer 17. The optical lens 11 and the image sensor 12 serve as an imaging unit. The AD converter 13 and the image processing DSP 14 serve as image processing means for converting an analog signal output from the imaging means into a digital signal, performing predetermined image processing on the digital signal, and generating a digital video signal. The AD converter 13 converts the analog image signal output from the image sensor 12 into a digital image signal. In addition, the image processing DSP 15 performs predetermined image processing on the digital signal output from the AD converter 13 to generate and output a digital video signal, and also output a horizontal / vertical synchronization signal and a clock (CLK) signal. . The power driver circuit 16 is a circuit that drives the image sensor 12. The microcomputer 17 serving as a control unit controls the DSP 14 and the power supply driver circuit 16.

  The camera module 10 is connected to a data processing unit including a CPU 19 and a memory 20 via a device internal bus 18. An LCD (liquid crystal) display unit 21 is connected to the apparatus internal bus 18 as image display means. The data processing unit including the CPU 19 generates an image quality adjustment parameter based on the analysis result for recognizing the digital video signal and feeds it back to the camera module 10. The camera module 10 automatically corrects the image quality based on the image quality adjustment parameter, and transfers the corrected image to the data processing unit. After image quality adjustment (correction), the barcode reading engine process is executed by the CPU 19 as an image pattern recognition process based on the digital video signal.

  The embodiment of the present invention is also configured as follows. That is, the camera module 10 includes an image signal generating unit that generates an image signal by photographing, and a correcting unit that corrects the image signal from the image signal generating unit based on an image quality adjustment parameter from the image analyzing unit. . The data processing unit including the CPU 19 analyzes the image signal from the image signal generation unit, determines an image quality adjustment parameter based on the analysis result, and outputs the corrected image corrected by the correction unit. Pattern recognition means.

  FIG. 2 is a flowchart for explaining the processing of FIG. In FIG. 2, after the apparatus power is turned on in step S1, when the user starts barcode reading from a menu or the like (step S2), the camera is immediately activated (step S3), and the DSP 14 in the camera module 10 performs predetermined processing. Is set (step S4). After the image quality setting is completed, the camera can shoot in step S4, and the video signal to be read is taken into the memory 20 via the device internal bus 18 (step S6).

  In the next step S7, the image data is read from the memory 20 into the reading engine of the CPU 19, and the image data is analyzed. This analysis includes evaluation of resolution, contrast, etc. by histogram analysis, etc., and it is judged whether or not this evaluation value has reached a level necessary for the engine to read the code (such as a target value derived from experience). .

  As a result of the image analysis in step S7, it is determined in step S8 whether correction is necessary. Here, if there is an item determined to require image quality correction (YES), the CPU 19 transmits a setting change instruction to the DSP 14 of the camera module 10 (step S9). Based on an instruction from the CPU 19, the DSP 14 of the camera module 10 executes image quality setting again under the control of the microcomputer 17, and the above series of shooting operations is executed. The image data taken into the memory 20 by photographing based on the new image quality setting is analyzed again in step S7, and whether or not image quality correction is possible is determined in step S8.

  The image quality correction and re-shooting operations are repeated until it is determined in step 7 that image quality correction is not necessary. Finally, after it is determined that image quality correction is not necessary, the barcode reading operation in the next step S10 is performed. Transition.

  In the next step S11, the CPU 19 performs a binarization process on the data based on the luminance information of the image data input by the recognition engine, and reads the information stored in the barcode. It is determined in step S11 whether or not the reading operation has been performed without any problem. If the reading is OK, the CPU 19 displays the result of step S12 and ends the series of operations (step S15). If the reading operation is NG in step S11, it is determined whether or not a retry is necessary (YES / NO) in step S13. If the number of retries is equal to or smaller than the predetermined number of retries, the process returns to step S5, Are repeated a predetermined number of times (retries). If the reading is NG even after the predetermined number of retries, the reading operation is abandoned in step 13, the unsuccessful reading result is displayed in step S14, and the operation is terminated (step S15).

  Through the operation described above, based on the analysis result of the target image (code) captured by the camera, the image quality adjustment parameter judged to be necessary for improving the recognition rate is fed back to the camera and read. Since the optimum image quality adjustment can be automatically performed, it is possible to improve the barcode recognition rate. Further, since the image quality adjustment can be automatically performed, it is possible to eliminate the need for securing the memory required during the processing.

  The configuration of this embodiment is the same as that of FIG. 1, and the operation thereof is also the same as the flow of FIG. In FIG. 1, an image sensor 12 of a camera module 10 converts an optical subject image from a lens 11 into an electrical signal. The image processing DSP 14 performs predetermined image processing such as color interpolation, color correction, and image quality adjustment on the digital signal output from the AD converter 13 to generate and output a digital video signal, and at the same time, horizontal and vertical synchronization. Signal and CLK signal are output. The I / F circuit 15 exchanges an output signal of the image processing DSP 14 and a control signal from a CPU 19 described later via the apparatus internal bus 18.

Here, the image analysis method in step S7 will be described. First, the resolution will be described with reference to the schematic diagram of the image in FIG. 3 and its output signal. Regarding the resolution, a second derivative operation is performed on the body image B in the obtained image G to extract a contour signal. As an example, FIG. 3 shows a video signal of the nth line and its secondary differential signal (contour signal), and a video signal of the mth line and its secondary differential signal (contour signal). The video signal level of the nth line is a, the level of the secondary differential signal is a, the video signal level of the mth line is b, and the level of the secondary differential signal is β. Here, the ratio between the video signal level and the second-order differential signal level of each line is α / a = kn and β / b = km, respectively, and the reference value of the ratio is k. At this time, the determination of the resolution is based on whether or not kn ≧ k and km ≧ k. When kn <k and km <k, it is determined that the predetermined resolution is not obtained. If it is determined that the resolution is insufficient based on the determination result, in step S9, the CPU 19 instructs the camera module 10 to increase the edge enhancement level so that km ≧ k and kn ≧ k.
The resolution of the captured image can be increased by an instruction to increase the contour enhancement level by the resolution determination, and as a result, the code recognition rate can be increased.

  Next, a contrast determination / adjustment method will be described with reference to FIGS. 4A and 4B are a schematic diagram of a screen and a contrast level diagram of FIG. 4A shows an example of a photographed image before contrast correction of an image with low contrast, and FIG. 4B shows an example of an image after contrast correction. 4A and 4B, “AA ′ column signal waveform” is a schematic waveform representing the signal waveform of AA ′ column in an analog manner, and “binary expected waveform” is “AA ′ column signal”. It is an expected schematic waveform when binarizing “waveform” based on “binarization threshold”. FIG. 5 schematically shows the histogram of the image before contrast correction and the histogram of the image after contrast correction. In the figure, the vertical axis represents quantity and frequency, the horizontal axis represents signal strength (in the case of 8 bits), the right end is white, and the left end is black.

  The histogram distribution of the low-contrast image shown in FIG. 4A is generally smooth like the level a in FIG. 5, and the distribution peak is not clear and many intermediate level signals are included. On the other hand, in the image with high contrast as shown in FIG. 4 (b), the sharpness of the black-and-white pattern increases, so the distribution peak is clearly poled in the histogram (level b in FIG. 5). The signal distribution at the intermediate level is reduced. Thus, the CPU 19 can control the adjustment parameter on the camera side so as to approach the ideal histogram distribution by analyzing the histogram distribution of the image data (step S9). As this adjustment parameter, it is conceivable to change the γ correction curve or increase the gain of the luminance signal.

  As shown in Fig. 4 (a), when an image with low contrast is binarized, the peak level of the video signal is low, so the threshold value is not reached, and the original code pattern cannot be reproduced accurately (Refer to the binarized expected waveform). As a result, the code contents cannot be read accurately. On the other hand, after the contrast correction as shown in FIG. 4B, the peak level of the video signal becomes high, the original code pattern can be accurately reproduced (binary expected waveform), and the code contents can be read.

  Through the operation described above, based on the analysis result of the target image (code) captured by the camera, the image quality adjustment parameter judged to be necessary for improving the recognition rate is fed back to the camera and read. Since the optimum image quality adjustment can be automatically performed, it is possible to improve the barcode recognition rate.

  In the present embodiment, the resolution and contrast have been described as the image quality adjustment parameters. However, the image quality adjustment parameters such as luminance may be added and analyzed for adjustment.

  FIG. 6 is a flowchart for explaining the second embodiment of the present invention. This will be described with reference to FIG. The difference between the first embodiment and this embodiment is that the feedback destination of the engine analysis result is not the DSP 14 of the camera module 10 but the CPU 19 mounted on the terminal device side.

  In this case, the engine analysis result is fed back to the image processing SW (software) executed by the CPU 19. That is, as the image processing SW, software for enhancing image quality is installed in a certain mobile phone, and in this embodiment, image quality is enhanced by using such software.

  In this embodiment, the data processing unit analyzes an image signal from the camera unit (10), determines an image quality adjustment parameter based on the analysis result, and outputs it, and an image quality adjustment parameter from the image analysis unit. The image processing means for correcting the image quality of the image signal so as to improve the image quality based on the predetermined image processing program, and the recognition means for recognizing the corrected image corrected by the image processing means.

  In this case, as in the first embodiment, the image analysis means analyzes the image signal, and determines and outputs an image quality adjustment parameter necessary for improving the recognition rate based on the analysis result. In this embodiment, the image processing unit corrects the image quality of the image signal so as to improve the image quality according to the predetermined image processing program by the image quality adjustment parameter from the image analysis unit, and the recognition unit corrects the corrected image corrected by the image processing unit. Recognize patterns.

  In the flow of FIG. 6, after the apparatus power is turned on (step S1), the user starts barcode reading from the menu or the like in step S2. In this case, the camera is immediately activated in step S3, and a predetermined image quality setting is made in the DSP 14 in the camera module 10 in step S4. After the image quality setting is completed, the camera can shoot in step S5, and the video signal to be read is fetched into the memory 20 via the apparatus internal bus 18 in step S6.

  Next, the image data taken into the memory 20 is read and subjected to predetermined video signal processing in step S7a by the image processing SW. The video signal process executed here is a process executed in a complementary manner to improve the image quality of the image output from the camera module 10 and is executed in the microcomputer 17.

  In the next step S7b, the image data is read from the memory 20 into the reading engine of the CPU 19, and the image data is analyzed. This analysis includes evaluation of resolution, contrast, etc. by histogram analysis. This evaluation value is judged by whether or not the engine has reached a level necessary for reading the code (such as a target value derived from experience). . As a result of the image analysis in step S7b, it is determined in step S8 whether image quality correction is necessary. If there is an item that is determined to require image quality correction (YES), a setting change instruction is transmitted from the CPU 19 to the image processing SW in step S9a. In step S7a, the image processing SW performs image quality setting again on the basis of an instruction from the CPU 19, and the series of photographing operations is performed. The image data taken into the memory 20 by photographing based on the new image quality setting is analyzed again in step S7b, and whether or not image quality correction is possible is determined in step S8.

  In the process up to this point, image quality correction and re-shooting operations are repeated until it is determined in step S8 that image quality correction is not necessary. After it is finally determined that image quality correction is not necessary, the reading operation in next step S10 is performed. Migrate to

  In step S10, the CPU 19 performs a binarization process on the data based on the luminance information of the image data input by the recognition engine, and reads the information stored in the barcode. In the next step S11, it is determined whether or not the reading operation has been performed without any problem. If the reading is OK, the result display (step S12) is performed and the series of operations is terminated (step S15). If the reading operation is NG in step S11, the CPU 19 determines whether or not a retry is necessary in step S13. If the number of retries is less than or equal to the predetermined number of retries, the process returns to the shooting operation in step S5, and a series of shooting, image data capture, and reading operations are repeated a predetermined number of times (retry). If the reading is NG even after the predetermined number of retries, the reading operation is abandoned in step S13, the reading unsuccessful result is displayed (step S14), and the operation is ended (step S15).

  The effects of the present embodiment described above are the same as those of the first embodiment described above, but communication with the camera module 10 is not required, so that the processing can be simplified and the time can be reduced.

  It should be noted that the barcode reading apparatus according to the present embodiment described above can be applied not only to a mobile phone but also to all electronic devices having a configuration using the camera module 10 shown in FIG.

  In addition, although a two-dimensional barcode has been described as an example of a reading target, it is obvious that the present invention can be applied to other reading targets, for example, a one-dimensional barcode such as a JAN code, and a reading device for character information.

  In the present embodiment, the case where the camera module 10 and the data processing unit are connected via the internal bus 18 has been described. However, the camera module 10 and the data processing unit may be connected by radio waves or infrared rays.

  The present invention can be applied not only to camera-equipped mobile terminals and mobile phones that can read barcodes, but also to various barcode readers, and can also be applied to general pattern recognition devices. The present invention can also be applied to equipped electronic devices.

It is a block diagram of a camera module explaining a 1st embodiment of the present invention. It is a flowchart explaining the process of FIG. It is a signal level diagram explaining the image level of FIG. (A) and (b) are signal level diagrams for explaining other image levels in FIG. It is a signal level diagram explaining the other image level of FIG. It is a flowchart explaining the 2nd Example of this invention. It is a block diagram explaining the image reading apparatus of a prior art example. It is a flowchart explaining the process of FIG. It is a block diagram explaining the diagnosis assistance apparatus of another prior art example. It is a flowchart explaining the process of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Camera module 11 Lens 12 Image pick-up element 13 AD converter 14 Image processing DSP
15 I / F (interface) circuit 15a External I / F
16 Power Driver Circuit 17 Microcomputer 18 Internal Bus 18a Bus 19 CPU
20, 20a Memory 20b Storage unit 20c Temporary storage unit 21, 21a, 21b LCD (display unit)
Reference Signs List 31 control unit 32 state detection unit 33 operation control unit 34, 34a image input unit 35 image processing unit 36 image compression unit 37 operation unit 41 input unit 42 processing unit 51 image data characteristic detection unit 52 threshold correction unit 53 diagnostic unit 55 Basic program 56 Image data characteristic detection program 57 Threshold value 58 Threshold value correction table

Claims (15)

Image signal generating means for generating an image signal by photographing, image analyzing means for analyzing the image signal from the image signal generating means, determining and outputting an image quality adjustment parameter based on the analysis result, and output from the image analyzing means A pattern recognition apparatus comprising: a correction unit that corrects the image signal from the image signal generation unit based on an image quality adjustment parameter; and a recognition unit that recognizes a pattern of the corrected image corrected by the correction unit. . Image pickup means, image processing means for converting an analog signal output from the image pickup means into a digital signal, performing predetermined image processing on the digital signal to generate a digital video signal, and control for controlling the image processing means A camera unit including means,
A data processing unit connected to the camera unit and executing an image pattern recognition process based on the digital video signal;
The data processing unit generates an image quality adjustment parameter based on an analysis result for recognizing the digital video signal and feeds it back to the camera unit. Further, the camera unit automatically generates an image based on the image quality adjustment parameter. A pattern recognition apparatus characterized by correcting image quality. Image signal generating means for generating an image signal by photographing, image analyzing means for analyzing the image signal from the image signal generating means, determining and outputting an image quality adjustment parameter based on the analysis result, and output from the image analyzing means An image processing unit that corrects the image quality of the image signal so as to improve the image quality by a predetermined image processing program based on an image quality adjustment parameter, and a recognition unit that recognizes a pattern of the corrected image corrected by the image processing unit. A pattern recognition device. 4. The pattern recognition apparatus according to claim 1, wherein the image quality adjustment parameter is image resolution or contrast. 5. The pattern recognition apparatus according to claim 4, wherein the image resolution is corrected by performing a second order differential operation on the input image signal to extract a contour signal and controlling the contour signal level to be higher than the video signal level. 5. The pattern recognition apparatus according to claim 4, wherein the image contrast is corrected by analyzing the histogram distribution for the signal level of the input image signal and controlling the histogram distribution to be a steep pattern. The pattern recognition apparatus according to claim 1, wherein the pattern recognition apparatus is a bar code reader. An electronic apparatus comprising the pattern recognition device according to claim 1. The electronic device according to claim 8, wherein the electronic device is a mobile terminal with a camera capable of reading a barcode. The image signal from the image signal generation means is analyzed by the image analysis means, and an image quality adjustment parameter is determined and output based on the analysis result. Based on the image quality adjustment parameter from the image analysis means, the image signal generation means A pattern recognition method characterized in that the image signal is corrected by a correction means, and the corrected corrected image is recognized by a recognition means. The image signal from the image signal generation means is analyzed by the image analysis means, and an image quality adjustment parameter is determined and output based on the analysis result, and the image processing means uses the image quality adjustment parameter from the image analysis means. A pattern recognition method, wherein the image signal is subjected to image quality correction so as to improve image quality by a predetermined image processing program, and the corrected image corrected by the image processing means is recognized by the recognition means. 12. The pattern recognition method according to claim 10, wherein the image quality adjustment parameter is image resolution or contrast. 13. The pattern recognition method according to claim 12, wherein the image resolution is corrected by performing a second derivative operation on the input image signal to extract a contour signal, and controlling the contour signal level to be higher than the video signal level. 13. The pattern recognition method according to claim 12, wherein the image contrast is corrected by analyzing the histogram distribution for the signal level of the input image signal and controlling the histogram distribution to be a steep pattern. 15. The pattern recognition method according to claim 10, wherein the pattern recognition method is a barcode reading method.

Images