A kind of measurement techniques for quality detection of meat detection system and method based on ultrasonic signal
Technical field
The invention belongs to ultrasonoscopy processing technology field, relate in particular to a kind of measurement techniques for quality detection of meat detection system and method based on ultrasonic signal.
Background technology
Meat and meat products thereof can provide protein, fat, inorganic salts and the vitamin etc. of needed by human body, are the important sources that the mankind obtain nutriment.Along with the continuous progress of society and economic fast development, people's living standard improves constantly, and consumption demand is growing, more and more pays close attention to the health of diet, and the requirement of measurement techniques for quality detection of meat is improved constantly.At present for the evaluation of measurement techniques for quality detection of meat, mostly be traditional subjective sensory evaluation or conventional instrument mensuration etc., subjective sensory evaluation mainly depends on the subjective factors such as people's vision, experience, lacks objectivity, evaluates precision low.Although conventional instrument has higher evaluation accuracy, apparatus system is all comparatively loaded down with trivial details, and operating process is complicated, destructive large, cannot obtain in real time, easily measurement techniques for quality detection of meat evaluate parameter; Number of patent application is the system and method for the fresh Meat of a kind of Fast Evaluation of 201110228938.5 proposition, and this system is comprised of imaging spectrometer, CCD digital camera, halogen tungsten lamp dc point light source, light-source stabilizer, camera controller, computing machine and image pick-up card.The method gathers the high spectrum image on pork surface, then, according to the physical and chemical parameter evaluation classification of measuring, finally adopts mathematical method to analyze the forecast model that high spectrum image is set up evaluation pork freshness.The method is less with respect to its destructiveness of conventional instrument, but equipment is various, system complex, operability is lower, convenience is poor.
The textural characteristics that meat muscle fibre presents is an important parameter of assessment measurement techniques for quality detection of meat quality, the orderly meat of clean mark is good than the mixed and disorderly unordered mouthfeel of texture, be of high nutritive value, and therefore how extracting quickly and accurately meat textural characteristics seems particularly important.
Summary of the invention
The invention provides a kind of measurement techniques for quality detection of meat detection system and method based on ultrasonic signal, be intended to solve that existing measurement techniques for quality detection of meat detecting instrument system is loaded down with trivial details, complicated operation, the poor technical matters of convenience.
Technical scheme provided by the invention is: a kind of measurement techniques for quality detection of meat detection system based on ultrasonic signal, comprise acquiring ultrasound image device, texture extraction module and image sampling module, described acquiring ultrasound image device is for gathering the ultrasonoscopy of meat to be measured, described texture extraction module obtains texture image for extract the textural characteristics of ultrasonoscopy by texture extraction algorithm, described image sampling module is for calculating pixel average and the pixel variance yields of described texture image, according to the texture uniformity coefficient of the pixel variance yields judgement texture image of texture image.
Technical scheme of the present invention also comprises: described acquiring ultrasound image device comprises ultrasonic probe, display and control panel, described ultrasonic probe is for gathering the ultrasonoscopy of meat diverse location to be measured, and the ultrasonoscopy collecting by the demonstration of described control panel control display.
Technical scheme of the present invention also comprises: also comprise image processing module, described image processing module is for carrying out cutting to the ultrasonoscopy collecting, retain the ultrasonoscopy in the region that comprises the main texture information of meat, and the ultrasonoscopy in this region is carried out to filtering, filtering noise.
Technical scheme of the present invention also comprises: the texture extraction algorithm that described texture extraction module adopts is specially: adopt two-dimensional empirical mode decomposition to do multiple dimensioned decomposition to ultrasonoscopy, obtain the ground floor two dimension intrinsic mode function that marginal information is the abundantest, and resolve the two-dimensional spectrum feature that two-dimentional intrinsic mode function extracts ultrasonoscopy, obtain the textural characteristics of ultrasonoscopy.
Technical scheme of the present invention also comprises: pixel average and pixel variance yields that described image sampling module is calculated described texture image specifically comprise: the sub-block that texture image sampling is divided into the different pixels of some, and add up the pixel value of each sub-block, calculate the pixel average of each sub-block, obtain pixel average and the pixel variance yields of entire image, the texture uniformity coefficient of the pixel variance yields judgement texture image by entire image, and using the texture uniformity coefficient of texture image as the index of evaluating measurement techniques for quality detection of meat.
Another technical scheme provided by the invention is: a kind of measurement techniques for quality detection of meat detection method based on ultrasonic signal, comprising:
Step a: gather the ultrasonoscopy of meat to be measured and show;
Step b: the textural characteristics that extracts ultrasonoscopy by texture extraction algorithm obtains texture image;
Step c: calculate pixel average and the pixel variance yields of described texture image, according to the texture uniformity coefficient of the pixel variance yields judgement texture image of texture image.
Technical scheme of the present invention also comprises: described step a also comprises: the ultrasonoscopy collecting is carried out to cutting, retain the ultrasonoscopy in the region that comprises the main texture information of meat, and the ultrasonoscopy in this region is carried out to filtering, filtering noise.
Technical scheme of the present invention also comprises: in described step b, the described texture image that obtains ultrasonoscopy by texture extraction algorithm specifically comprises: adopt two-dimensional empirical mode decomposition to do multiple dimensioned decomposition to ultrasonoscopy, obtain the ground floor two dimension intrinsic mode function that marginal information is the abundantest, and resolve the two-dimensional spectrum feature that two-dimentional intrinsic mode function extracts ultrasonoscopy, obtain the textural characteristics of ultrasonoscopy.
Technical scheme of the present invention also comprises: in described step c, pixel average and the pixel variance yields of the described texture image of described calculating are specially: the sub-block that texture image sampling is divided into the different pixels of some, and add up the pixel value of each sub-block, calculate the pixel average of each sub-block, obtain pixel average and the pixel variance of entire image, the texture uniformity coefficient of the pixel variance yields judgement texture image by entire image, and using the texture uniformity coefficient of texture image as the index of evaluating measurement techniques for quality detection of meat.
Technical scheme of the present invention also comprises: after described step c, also comprise: set pixel variance threshold values, and whether the pixel variance that judges described texture image is less than described pixel variance threshold values, if be less than pixel variance threshold values, represent that tested measurement techniques for quality detection of meat is higher, if be greater than pixel variance threshold values, represent that tested measurement techniques for quality detection of meat is lower.
Technical scheme tool of the present invention has the following advantages or beneficial effect: the measurement techniques for quality detection of meat detection system based on ultrasonic signal of the embodiment of the present invention and method are by adopting two-dimensional empirical mode decomposition and two-dimensional spectrum feature, obtain texture image, according to the pixel variance yields of texture image, judge the texture uniformity coefficient of this image, and using the texture uniformity coefficient of image as the index of evaluating measurement techniques for quality detection of meat, system is simple, convenient operation, can react measurement techniques for quality detection of meat intuitively, fast and accurately.
Accompanying drawing explanation
Accompanying drawing 1 is the structural representation of the measurement techniques for quality detection of meat detection system based on ultrasonic signal of the embodiment of the present invention;
Accompanying drawing 2 is process flow diagrams of the measurement techniques for quality detection of meat detection method based on ultrasonic signal of the embodiment of the present invention.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
Referring to Fig. 1, is the structural representation of the measurement techniques for quality detection of meat detection system based on ultrasonic signal of the embodiment of the present invention.The measurement techniques for quality detection of meat detection system based on ultrasonic signal of the embodiment of the present invention comprises acquiring ultrasound image device, image processing module, texture extraction module and image sampling module, wherein,
Acquiring ultrasound image device comprises ultrasonic probe, display and control panel, and ultrasonic probe is used for gathering the ultrasonoscopy of meat diverse location to be measured, and controls by control panel the ultrasonoscopy that display demonstration collects; Wherein, the mode that ultrasonic probe gathers ultrasonoscopy is: the long axis direction of ultrasonic probe is vertically positioned over to the surface of meat to be measured, and by adjusting the position of ultrasonic probe, gathers the meat ultrasonoscopy of diverse location.
Image processing module, for the ultrasonoscopy collecting is carried out to cutting, retains the ultrasonoscopy in the region that comprises the main texture information of meat, and the ultrasonoscopy in this region is carried out to filtering, filtering noise;
Texture extraction module is for obtaining the texture image of ultrasonoscopy by texture extraction algorithm, texture extraction algorithm is specially: adopt two-dimensional empirical mode decomposition to do multiple dimensioned decomposition to ultrasonoscopy, obtain the ground floor two dimension intrinsic mode function that marginal information is the abundantest, and resolve the two-dimensional spectrum feature that two-dimentional intrinsic mode function extracts ultrasonoscopy, obtain the textural characteristics of ultrasonoscopy;
Image sampling module is for being divided into texture image sampling the sub-block of the different pixels of some, and add up the pixel value of each sub-block, calculate the pixel average of each sub-block, obtain pixel average and the pixel variance (Variance of entire image, the difference of each data and average square average) value, the texture uniformity coefficient of the pixel variance yields judgement texture image by entire image, and using the texture uniformity coefficient of texture image as the index of evaluating measurement techniques for quality detection of meat, pixel value variance can be reacted the uniformity coefficient of image texture preferably, the image pixel variance that texture is uniformly distributed (meat that quality is high) is less, on the contrary, the image pixel variance of texture disorderly and unsystematic (meat that quality is lower) is larger, in another embodiment of the present invention, also can be by setting pixel variance threshold values as the index of evaluating measurement techniques for quality detection of meat, for example, setting pixel variance threshold values is 0.5, to be less than 0.5 measurement techniques for quality detection of meat higher for pixel variance, it is lower that pixel variance is greater than 0.5 measurement techniques for quality detection of meat, by calculating the pixel distribution variance of texture image, quantize texture uniformity coefficient, simple to operate, and can be directly perceived, fast, react accurately measurement techniques for quality detection of meat, wherein, the cutting apart quantity or partitioning scheme and can set with actual demand of texture image, in embodiments of the present invention, the sub-block pixel of cutting apart is 10 * 10.
Referring to Fig. 2, is the process flow diagram of the measurement techniques for quality detection of meat detection method based on ultrasonic signal of the embodiment of the present invention.The measurement techniques for quality detection of meat detection method based on ultrasonic signal of the embodiment of the present invention comprises:
Step 100: gather the ultrasonoscopy of meat diverse location to be measured, and show the ultrasonoscopy collecting;
In step 100, the long axis direction of ultrasonic probe is vertically positioned over to the surface of meat to be measured, and by adjusting the position of ultrasonic probe, gathers the meat ultrasonoscopy of diverse location.
Step 200: the ultrasonoscopy collecting is carried out to cutting, retain the ultrasonoscopy in the region that comprises the main texture information of meat, and the ultrasonoscopy in this region is carried out to filtering, filtering noise;
Step 300: the texture image that obtains ultrasonoscopy by texture extraction algorithm;
In step 300, texture extraction algorithm is specially: adopt two-dimensional empirical mode decomposition to do multiple dimensioned decomposition to ultrasonoscopy, obtain the ground floor two dimension intrinsic mode function that marginal information is the abundantest, and resolve the two-dimensional spectrum feature that two-dimentional intrinsic mode function extracts ultrasonoscopy, obtain the textural characteristics of ultrasonoscopy.
Step 400: texture image sampling is divided into the sub-block of the different pixels of some, and adds up the pixel value of each sub-block, calculate the pixel average of each sub-block, obtain pixel average and the pixel variance of entire image;
In step 400, the cutting apart quantity or partitioning scheme and can set with actual demand of texture image, in embodiments of the present invention, the sub-block pixel of cutting apart is 10 * 10.
Step 500: the texture uniformity coefficient of the pixel variance yields judgement texture image by entire image, and using the texture uniformity coefficient of texture image as the index of evaluating measurement techniques for quality detection of meat;
In step 500, pixel value variance can be reacted the uniformity coefficient of image texture preferably, the image pixel variance that texture is uniformly distributed (meat that quality is high) is less, on the contrary, the image pixel variance of texture disorderly and unsystematic (meat that quality is lower) is larger, in another embodiment of the present invention, also can be by setting pixel variance threshold values as the index of evaluating measurement techniques for quality detection of meat, for example, setting pixel variance threshold values is 0.5, to be less than 0.5 measurement techniques for quality detection of meat higher for pixel variance, and it is lower that pixel variance is greater than 0.5 measurement techniques for quality detection of meat;
Technical scheme tool of the present invention has the following advantages or beneficial effect: the measurement techniques for quality detection of meat detection system based on ultrasonic signal of the embodiment of the present invention and method are by adopting two-dimensional empirical mode decomposition and two-dimensional spectrum feature, obtain texture image, according to the pixel variance yields of texture image, judge the texture uniformity coefficient of this image, and using the texture uniformity coefficient of image as the index of evaluating measurement techniques for quality detection of meat, system is simple, convenient operation, can react measurement techniques for quality detection of meat intuitively, fast and accurately.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.