TW201044285A - Apply video image recognition function to search specific objects - Google Patents

Abstract

The method of applying video image recognition function to search specific objects.

Description

201044285 IV. Designated representative map: (1) The representative representative of the case is: Figure 6 is the picture. (2) A brief description of the symbol of the representative figure: 130: Target object to be searched 155 • Space related to the person most relevant to the disappearance of the object Video screen 160: Person looking for object 0 17〇: Image recognition object search system 5. If there is a chemical formula in this case, please disclose the best way to display the characteristics of the invention. 6. Description of the invention: [Technical field of the invention] The present invention relates to a method for processing image data, using image recognition to help Look for objects. # [Prior Art] It is currently available to view all recorded images to find a specific object, which is labor intensive.呀 [The content of the invention] Or not, this hair duck ride _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Recording images to help people in the room find the location of a particular object, such as when people forget to place the remote control of the TV, they can access the recorded image to find out where the last person touched the TV's remote control. In turn, it helps people narrow down the search range and can find the remote control of the TV. Or people sometimes put them in their hands; when they leave the car key and forget where to place the car keys, they can read the recorded images to find out where people dare to touch the car keys, thus helping people narrow down the search range. The car key can finally be found. 〇 [Embodiment] Several sets of cameras are installed in several places in the room (Fig. 1). σ The camera image is divided into people and no one (Fig. 2). - Generally speaking, the movement of an object depends on a person to take or move it. Someone's video clip is used to distinguish whether the person has the object (Fig. 3. 〇3 the position of the object or find out that the object finally appears. The fragment image is used to narrow the search range (Fig. 4). The person uses this small range to do the search for the object (Fig. 5). [Simplified illustration] Fig. 1: Putting a few _ shadows in the room _ Several places Figure 2: Recording the image of the camera and distinguishing the image into three. Figure 3: Accessing the image of a person to divide it. Figure 4: Finding the woman’s poem Image to narrow the search scope 201044285 Figure 5: People use this small range to find _ to find the object Figure 6. Use the image identification button to find a specific object Figure 7: _ tear the Wei people's (4) objects The service of the possible location is out of a special [main component symbol description] 100: indoor configuration diagram no: surveillance camera 120. A person appears in the image plane 130: the target object to be searched 140: the object disappears most There are related characters 15 〇. With the disappearance of objects most _ The character's smoke space video check 155: the object with the disappearance of the object _ the empty _^ face 1 160: the person looking for the object ~ face 2 170: image recognition object search system 180: the object or operator to find the object 19〇. Provided by the trustee or operator of the object search service 200: Recorded image data 210: Internet or Internet

Claims (1)

201044285 VII. Patent application scope: 1. A method that can apply image recognition function to find specific objects. 2. A hardware device that can apply image recognition to find specific objects. 3. In combination with the above, the image recognition function is applied to the service of finding a specific object. 5