Sapera Processing 5.0 TM Image analysis and processing database T E C H N I C A L S P E C S Key features Features •UseMMX, SSE,SSE2Optimized to improve performance •Multi-thread and multi-processor support Quick Development •High-levelC ++Class to simplify the development process Flexibility •Hardware independence for maximum flexibility Measurable •Special Tool application, mainly to complete the search,OCR, I-D, 2-DBar code decoding andBlobAnalysis Sapera Processing 5 is a Windows-based image library consisting of a complete set of image Processing and analysis function modules.
Sapera Processing 5 adopts MMX, SSE (Streaming SIMD Extensions) and SSE2
To meet the challenges of the current image system. Sapera Processing 5 provides image Processing, search, OCR, bar code decoding, and blob
Analysis and other tools, and provides a scalable design. Each tool provides separate modules to maximize your budget. Sapera Processing 5 is based on high-level
C ++ class to simplify the visual development process. Software Structure Image processing tools This tool includes a series of highly optimized basic image processing function modules. Filter • Generalized one-and two-dimensional convolution • Low-pass, high-pass, and two-dimensional Filters • Laplace transformation • Gaussian Transformation • Range: 8-way edge enhancement • Gradient: Amplitude and Phase • Edge and Line Detection Ry • Flip: horizontal, vertical, transpose, and 90-degree rotation • Rotate at any angle • Conversion: Translation and rotation with integer or secondary Pixel Precision • Zoom with image size or proportional factor • Cropping: horizontal or vertical • Affine and fluoroscopy calibration and distortion • Polar Coordinate Conversion Measurement • Basic statistics, such as histogram based on image pixel values, pixel calculation, average value, variance, minimum value, and maximum value • Horizontal and vertical projection • Histogram processing: filtering and Peak Detection • Regular or unconventional interconnectivity • Degree of difference • Vector difference • Differential filter (differences between neighboring pixels per pixel) Morphology • Binary expansion, corrosion, opening, closure, and median processing of 1 or 8-bit Images • Grayscale expansion, corrosion, opening, closure, and median processing of 8 or 16-bit Images • Double expansion and corrosion of labeled images • Contour, skeleton, dilution and thickening Filters • Conditional binary expansion • Morphological gradient • Hierarchical Filter • Hit-or-miss Conversion • Top-hat Conversion Point-to-Point • Operations on one or two frames: addition, subtraction, multiplication, division, absolute difference, minimum value, maximum value, and user-defined expression • Logical operations for one or two frames of images: And, Or, exclusive or • Fixed or adaptive thresholds • Gaussian Noise Generation • Query table conversion • Manual contrast balancing or histogram-based segmentation Split • Distance graph • Edge Detection • K-means clustering • Local Peak Detection • Regional Growth • Watershed Transformation • Zero-intersection detection Conversion • Spectral conversion: FFT, DCT, and DHT) • Hough conversion: direct or reverse Conversion Search Tool (pattern matching) This software is an advanced pattern matching engine designed for processing images collected under weak uneven lighting conditions. The fast, powerful, and accurate algorithms of this software are suitable for the demanding applications of the semiconductor and electronics manufacturing industries. General Features • User-selectable algorithms (region-based or edge-based) • Adjustable rectangular ROI • Multi-Model Training • Adjustable rotation and scaling ratio Region-based algorithm features • NCC-based algorithms • High-speed processing of conventional Images • Almost no parameters required • High Noise • Secondary pixel precision: 1/64> pixels Edge-based algorithm features • Target contour-based algorithms • Rotate, scale, and contrast independently • Allow large occlusion • Large contrast Conversion • Large overlaps and touching targets • User-selectable model outlines • Precision of sub-pixel: 1/40 pixels OCR Tool (Optical feature recognition) This software is a complete optical feature recognition toolkit with proportional and contrast immutability. It supports 3D and dot matrix fonts and can be viewed as user fonts. To achieve high speed and accuracy, its powerful recognition technology has been optimized. The OCR tool's ability to process highly degraded images makes it an ideal choice for pharmaceutical, electronic, and semiconductor tool applications. Features • User-trainable (more than one character at a time) • Three-dimensional and dot matrix fonts (tilting, tilting fonts with Support Technology) • Proportional immutability (up to 400%) • Allow non-linear background changes • Support for touching feature Training • Optimization: three different algorithms and three resolutions • Complete Morphological Processing • Domain: limits the range of acceptable features at specific locations • Time-based execution BlobTools(BlobAnalysis) This software is designed to separate targets from the background and calculate a series of geometric and gray-scale characteristics. It is very suitable for the detection of electronic and semiconductor industries. Features • Supports 1-bit, 8-bit, 16-bit, binarization or grayscale input images • Different Levels of input (grayscale images, threshold (value) images, or RLE) are allowed to interface with hardware accelerators for threshold value and/or RLE • Computing of over 50 space and grayscale blob features, including: the minimum boundary, size, orientation, or more of the area, center, perimeter, number of holes, extended line, circle, elliptical, convex area, and box position. • Sort and select blobs Based on Computing Features • Feature groups selected for Calculation • Automatically set the threshold value for the input grayscale image • Optical input RLE (Run-Length-Encoding) allows users to define Computing Features Barcode tool (1-D and 2-D barcode) Bar Code tool is a series of functional modules used for one-dimensional and two-dimensional bar code decoding. It supports multiple standard codes in industrial applications. Bar Code tools are designed for degradation, low
Illumination image processing, which includes fast and powerful algorithms and functions. The image-based barcode tool is faster than laser-based barcode scanning. Such as OCR and Search
Bar Code tools provide more operation flexibility. Features • Pure digital Barcode EAN13 EAN8UPC-A UPC-E Line Scan 2 of 5 Bar Code (USD-4, NW-7 or 2 of 7) PostNet (3 of 5) RSS14 • Bar Code including text and numbers Code 39 (3 of 9) Code 128 BC412 • 2-D Barcode Data Matrix (ECC200) • General 1-D barcode allows any size and tiny rotation (excluding PostNet) The 2-D barcode allows full rotation and dimension variation of approx +/-50% Automatic one-dimensional Barcode type, orientation, Proportional Speed Optimization Algorithm Specify the region of interest when there is no surface (you may need to specify the speed when increasing) Optional area test features (only for some encoding) System Requirements • Pentium III higher CPU • MicrosoftWindows NT4.0, 2000, XP and XP Server • VGA/SVGA display • 64 MB system memory • 30 MB hardware drive space • Microsoft Visual C/C ++ 6.0 or higher |
