Continuous microbiotal recognition method

• US 6,130,956 A
• Filed: 02/17/1998
• Issued: 10/10/2000
• Est. Priority Date: 02/17/1998
• Status: Expired due to Fees

• Current Assignee: Francis M. Butterworth, Manohar Das
• Original Assignee: Francis M. Butterworth, Manohar Das
• Inventors: Das, Manohar, Butterworth, Francis M.
• Primary Examiner(s): Johns, Andrew W.
• Application Number:
  US09/024,184
• Time in Patent Office: 966 Days
• Field of Search: 382/100, 382/128, 382/133, 382/155, 382/190, 382/226, 382/275, 348/79, 348/81, 356/39, 250/564, 250/565, 250/573, 250/576, 250/430, 250/431
• US Class Current: 382/100
• CPC Class Codes: G01N 15/147 the analysis being performed on a sample stream
• CPC Class Codes: G06V 20/69 Microscopic objects, e.g. biological cells or cellular parts

Abstract

A method to continuously identify and recognize microbiota in water comprises random sampling of a water source, directing the sample to an optical plane where microscopic CCD/video images are obtained and stored in analogue and digitized form. Statistical data on organism type and number will be simultaneously collected. Using novel software the digitized images are filtered and segmented. Segmented images will activate software to extract morphological and motion-related features. The extracted features are used to search a database for matching, specific microbiota profiles in a recognition and classification step. Software will enable classification, together with the statistical data will ascertain the biological status of the water. A flizzy-logic decision tree will trigger actuator/annunciator relays that manipulate water intake valves and send signals through modems for telephone warning messages.

Claims

1. A method of continuously recognizing microbiota in various types of water, including sewage, river and drinking water, comprising the steps of:

performing random sampling of a water source using an aquatic sampling apparatus;

passing the sampled water through a mesh filter and diverting a smaller sample of the water into a tube using a peristaltic pump;

conveying the smaller sample of water into a microscope viewing chamber and optical plane using said tube and said peristaltic pump;

capturing images in said optical plane using a CCD/video camera;

wherein said images are recorded for analogue storage;

digitizing said images for processing and chronological storage;

filtering the digitized images to remove extraneous objects and noise using edge-preserving image restoration and enhancement software;

segmenting the filtered images;

extracting morphological-, color-, and motion-related features from the segmented images;

classifying microbiota in the images by searching a database in a hierarchical fashion using the extracted features to match specific microbiotal class profiles;

constructing estimated cell concentration profiles for different microbiotal classes using the classification results obtained from N consecutive image frames;

monitoring the estimated cell concentration profiles at regular time intervals using a rule-based expert system to scan a fuzzy decision tree; and

generating appropriate command and control signals and transmitting these signals to alarm and actuating devices by means of cables and local area network or modem connections.