Datum Data Co., Ltd: Data Entry, Data Capture, Data Processing

Datum Data Successfully Attended 10th Hong Kong International ICT EXPO

Chinese

	Data Entry

	Data Processing

	Data Conversion

	Document Management Sysiem(DMS)


	Software Development

Data processing ---Image processing

Image assessment

Raw images, including those that are already in electronic form, may come in the wrong orientations or out of order, contain dirt, or be distorted in some fashion. They may even be damaged to a certain degree. Before they go through proper processing, these defects need to be thoroughly checked and a remedy plan needs to be devised and carried out. Datum engineers, armed with their experience and with software and hardware on hand, can quickly devise a strategy for processing a particular set of raw images. Based on the nature of a particular project, the initial assessment, and accuracy and quality requirements from clients, recommendations and work proposals can be developed with close consultation with clients. Datum has developed a series of patented award-winning software programs that are specifically tailored for this task.

Cleansing, Curation and Validation

Software based tools plus manual work will first locate and mark defects/noise in the images. Stained spots can be restored. Noise is removed. Contrast can be altered. Disorientation is corrected. Boundaries are well defined.This series of treatments will bring the images to a uniform format for further processing, be it OCR, visualization, or otherwise. Images may be validated against a carefully designed set of criteria to make sure they are in conformance with standards.

Change Detection

Change detection is an important technique and a vital step in OCR technology.Combining artificial intelligence technology, spatial location analysis, and comparisons with standard data (character) sets, characters can be delimited, separated and recognized. This process is highly sophisticated and complex.

2D and 3D Visualization

2D images are easier to generate and visualize as they are often captured in paper/film or screens. Difficulty comes when 3D visualization is desired from a set of 2D images. This requires expertise in computer vision, artificial intelligence and computer graphics. A computer model needs to be first devised based by physical simulation, scanning from a real-world object or procedural creation. Based on this model, 3D images can be rendered via a complex process considering many factors including, but not limited to, light transport, scattering and projection. This technology is popular in the health care industry such as in CT-scan visualization.