Also, since the correction values C1 to C12 of the twelve correctors 19B are obtained by solving the optimization problem that minimizes the sum of squares of the difference between the sum of the temperatures in each pixel of the second thermal image and the predetermined temperature, the optimum correction values C1 to C12 can be obtained by calculation without actually trying various combinations of correction values C1 to C12. In other words, the correction values C1 to C12 that minimize the evaluation function V represented by the equation (2) can be easily and efficiently obtained.

Further, when the first temperature distributions are obtained, the twelve LED arrays 400 in the area 401 are lighted one by one (individually), and therefore the correction values at the time of lighting in response to a sudden heat generation of the electronic device 500 in a limited heating zone can be obtained.

Also, the correction values C1 to C12 are set such that the correction values C3 to C12 for the ten correctors 19B corresponding to the ten LED arrays 400 around the presence area 401A within the area 401 are greater than the correction values Cl and C2 for the two correctors 19B corresponding to the LED arrays 400 within the presence area 401A. Therefore, adjustment can be performed to decrease the temperature of the portion where the temperature tends to be high, such as the central portion of the area 401, and to increase the temperature of the portion where the temperature tends to be low, such as the surroundings, thereby achieving uniformity of the temperature distribution within the area 401.