This paper presents a new texture representation called the bidirectional texture function (BTF), which captures the variation in texture with illumination and viewing direction. The authors describe a BTF database containing over 14,000 images from 61 different samples, each observed with over 200 different combinations of viewing and illumination directions. They also present a BRDF database with reflectance measurements for the same 60 samples. Both databases are publicly available and have important implications for computer graphics. The BTF is a texture image parameterized by illumination and viewing angles, and is analogous to the BRDF, which describes the reflectance of a surface. The BTF database can be used for development of 3-D texture algorithms, as traditional 2-D texture synthesis and mapping do not account for changes in texture appearance with viewing and illumination directions. The BTF database covers a diverse collection of rough surfaces and captures the variation of image texture with changing illumination and viewing directions. The authors describe a measurement procedure involving a robotic manipulator and CCD camera to allow simultaneous measurement of the BTF and BRDF of large samples. Radiometric calibration is performed to interpret pixel values as radiance, using a standard calibration card and sample calibration data. The calibration process involves determining the relationship between radiance and pixel values, and estimating the camera sensitivity curves for the red, green, and blue CCDs. The BRDF database provides a thorough investigation of the reflectance properties of real-world rough surfaces. This database fills a long-standing need for a benchmark to test and compare BRDF models. The authors also discuss the implications of their findings for computer graphics, emphasizing the importance of 3-D texture rendering algorithms and the need for accurate BRDF models. The databases are publicly available and have significant implications for computer graphics and computer vision.This paper presents a new texture representation called the bidirectional texture function (BTF), which captures the variation in texture with illumination and viewing direction. The authors describe a BTF database containing over 14,000 images from 61 different samples, each observed with over 200 different combinations of viewing and illumination directions. They also present a BRDF database with reflectance measurements for the same 60 samples. Both databases are publicly available and have important implications for computer graphics. The BTF is a texture image parameterized by illumination and viewing angles, and is analogous to the BRDF, which describes the reflectance of a surface. The BTF database can be used for development of 3-D texture algorithms, as traditional 2-D texture synthesis and mapping do not account for changes in texture appearance with viewing and illumination directions. The BTF database covers a diverse collection of rough surfaces and captures the variation of image texture with changing illumination and viewing directions. The authors describe a measurement procedure involving a robotic manipulator and CCD camera to allow simultaneous measurement of the BTF and BRDF of large samples. Radiometric calibration is performed to interpret pixel values as radiance, using a standard calibration card and sample calibration data. The calibration process involves determining the relationship between radiance and pixel values, and estimating the camera sensitivity curves for the red, green, and blue CCDs. The BRDF database provides a thorough investigation of the reflectance properties of real-world rough surfaces. This database fills a long-standing need for a benchmark to test and compare BRDF models. The authors also discuss the implications of their findings for computer graphics, emphasizing the importance of 3-D texture rendering algorithms and the need for accurate BRDF models. The databases are publicly available and have significant implications for computer graphics and computer vision.