This study is about determination of fire origin by using analysis of shadow that was recorded CCTV data at the fire scene. This analysis is based on straight and radiate nature of light. At fire experiment with about 1m flame, we conformed that 2-dimensional extension line is focused at the bottom of the fire. If the fire is burning at the same level with shadow, it indicate the point of origin exactly. In 3-dimensional analysis that connect extensional line between distinctive points the shadow and the objet, the line focused in the Ø 50cm-circle. We estimate the reason of that is because of the character of combustion of gases. The line indicate not the point of origin but the flame that is over the point of origin. thus, you have to consider the line indicate the flame when you do 3-dimensional analysis.

Computational algorithms and their implementations are studied for the treatments of kernel function’s shadow effects, which occurs in the application of the BEM(boundary element method) for radiation heat transfer analysis. In this problem, surfaces are assumed to be diffuse and gray placed in two-dimensional enclosures with transparent medium. Self-blocking or third party blocking is possible when the radiosity from the BEM nodes cannot reach the destination points. Also the third party blocking can be the partial or total blocking case. Algorithms, which can accurately recognize the each blocking cases and reflect the shadow effects to the BEM kernel function, are studied in this paper. Effective implementation methods are presented, and their results are verified by the test problem