会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 2. 发明专利
    • AREA DIVIDING METHOD
    • JPH02267675A
    • 1990-11-01
    • JP8790589
    • 1989-04-10
    • HITACHI LTD
    • MOTOIKE JUNHASHIZUME AKIHIDESUZUKI RYUICHI
    • G01N33/49G06T1/00
    • PURPOSE:To accurately and efficiently perform the area division of each constituent by dividing a background from a blood image, finding the density histogram of each image in an area where the nucleus and the cytoplasm of an erythrocyte and a leucocyte are mixed, and dividing the erythrocyte, the cytoplsm, and the nucleus by applying a threshold processing. CONSTITUTION:By using the blood images ((r) image, (g) image, and (b) image) of a red component, a green component, and a blue component inputted at every classification of wavelength by a photoelectric conversion, means, the background B is separated from other from the density historgram of the (g) image. Next, the erythrocyte, R, the cytoplasm C, and a threshold value to divide the cytoplasm C from the nucleus N are set from the density histogram of a differential image (g-b image) between the (g) image and the (b) image from which the background B is eliminated, and the division of the erythorocyte R from the cytoplasm C is performed by setting the threshold values of two differential images i.e. the (g-b) image and an (r-b) image, and a part which satisfies two conditions is set as the area of the erythrocyte R. In such a way, it is possible to accurately and efficiently perform the area division of the blood image.
    • 3. 发明专利
    • LEUKOCYTE CLASSIFYING DEVICE
    • JPH01147365A
    • 1989-06-09
    • JP30561087
    • 1987-12-04
    • HITACHI LTD
    • HASHIZUME AKIHIDEMOTOIKE JUNSUZUKI RYUICHI
    • G01N33/48
    • PURPOSE:To detect and classify eosinophil leukocytes having poor nucleus dyenability by determining the density width in a red wavelength region and the ratio between the density width in the red wavelength region and the density width in a green wavelength region. CONSTITUTION:The white light from a light source 41 is guided to a microscope 42 and the transmitted light of a blood coated specimen 43 set on the microscope is spectrally divided by an optical spectroscope 44 to 3 wavelength regions of red, green and blue. The spectrally divided light rays are converted by photoelectric transducers 451-453 to electric signals which are converted by A/D converters 461-463 to digital signals. The digital signals are stored in image memories 471-473. The threshold to extract the red cell region in the blood images is determined in accordance with the contents of the image memories. The density histogram is then analyzed by a characteristic extractor 48 to determine the density width 22 in the red wavelength region and the density width 23 in the green wavelength region. Whether the eosinophil leukocyte having the defective nucleus dyeability or not is classified from the position in space where the density width 22 and the ratio of the density width 22 and the density width 23 exist.
    • 5. 发明专利
    • PATTERN RECOGNITION PROCESSING METHOD
    • JPH08221380A
    • 1996-08-30
    • JP2148295
    • 1995-02-09
    • HITACHI LTD
    • MITSUYAMA SATOSHIMOTOIKE JUN
    • G01N33/493G06F15/18G06N3/00G06T7/00
    • PURPOSE: To easily retrieve an erroneously recognized pattern and to improve the efficiency of correction and re-recognition operations by selecting respectively corresponding classes for a prescribed number in the descending order of the output values of respective output elements and defining a recognized result. CONSTITUTION: Among the output elements of a neural network, a threshold value t1 for a difference between a maximum value and a second value and the threshold value t2 for the maximum value of the output element are set. Among the output values of the respective output elements, the difference between the maximum value p1 and the second value p2 is calculated and compared with the threshold value t1. When the difference between the p1 and the p2 is more than the t1, the p1 is compared with the threshold value t2, and when the p1 is smaller than the t2 or when the difference between the p1 and the p2 is smaller than the t1, the contents of data are displayed on a CRT screen. By comparing the stored and displayed maximum value p1 and second value p2 in such a manner, whether or not it is a pattern recognized result with the high possibility of erroneous recognition is easily discriminated and a different processing is performed again only for the pattern.
    • 6. 发明专利
    • FLOW TYPE PARTICLE IMAGE ANALYZING APPARATUS
    • JPH0815256A
    • 1996-01-19
    • JP14599194
    • 1994-06-28
    • HITACHI LTD
    • KOJIMA YASUAKIMOTOIKE JUNMITSUYAMA SATOSHIHASHIZUME AKIHIDE
    • G01N33/48G01N15/02G01N15/14
    • PURPOSE:To provide a flow-way particle image analyzing apparatus which is suitable to improve the efficiency of classifying process for particles, prevent the insufficiency of memory capacity for images, prevent a flow system from being choked, and improve the classifying precision and at the same time has good visibility of the particles. CONSTITUTION:Particles in a sample is dyed by mixing the sample and a dyeing reagent in a dyeing tank 34, the resulting sample is made to flow in a flow cell 100, the particles flowing in the flow cell 100 are detected by an image taking means 101 and a particle detecting means 103, and a particle image is produced. The produced image is analyzed by a particle analyzing means 102 and the particles are classified. A solution I, which is a solution containing 0.8X1/10 mole/l of Trypan blue and a solution II, which is a solution containing 1.0X1/10 mole/l Erythrosine, are mixed with a solvent, which is a buffer solution containing 1/15mole/l phosphoric acid (pH 6.8) in the solution I: the solution II=2:1 ratio and furthermore as a stabilizer, sodium azide is added to the solution in 0.1%, ratio, and after that the resulting solution is filtered and stored to be used for the dyeing reagent.
    • 10. 发明专利
    • CLINICAL INSPECTION SYSTEM
    • JPH11142411A
    • 1999-05-28
    • JP30250797
    • 1997-11-05
    • HITACHI LTD
    • MOTOIKE JUNMATSUOKA YOSHIOTAKAGI YOSHIMITSU
    • G01N35/00G06F19/00G06Q50/22
    • PROBLEM TO BE SOLVED: To conduct centralized control of data and secure the reliability of inspection data by connecting multiple inspection devices and data control devices arranged in inspection rooms with a network. SOLUTION: Multiple inspection devices and multiple data control devices 10-40 arranged in inspection rooms are connected by a network 70 such as a local area network(LAN). Data are centrally controlled, inspection methods can be referred from any of the data control devices 10-40, and a system with a high degree of freedom can be constructed. Inspection data from the data control devices 10-40 are transferred to a database server 50 through the network 70 and are stored in a memory device 60. Each screen of the data control devices 10-40 is constituted of a region controlled by order information and a region displaying the order information or specimen information, thereby inspection results can be accurately controlled, and the inspection data can be referred between inspection divisions.