Understand the working principle and function of the vision sensor in three minutes

In the machine vision industry, vision sensorsWith the ability to capture thousands of pixels of light from an entire image, the sensor can "see" a very detailed image of the target, regardless of the distance of meters or centimeters. After capturing the image, the vision sensor compares it to a benchmark image stored in memory for analysis. For example, if a vision sensor is set to identify a machine part with eight bolts inserted correctly, the sensor knows that it should reject parts with only seven bolts, or parts with misaligned bolts. In addition, the vision sensor can make a judgment regardless of where the machine component is located in the field of view, regardless of whether the part is rotated within a 360-degree range.

The role of vision sensors

Vision sensors are the direct source of information for the entire machine vision system, and are mainly composed of one or two pattern sensors, sometimes with light projectors and other auxiliary equipment. The primary function of a vision sensor is to acquire enough raw images to be processed by the machine vision system. Image sensors can be laser scanners, line and area CCD cameras, TV cameras, or emerging digital cameras.

Types of vision sensors

The application of vision sensors in intelligent networked vehicles is in the form of cameras, which are generally divided into monocular cameras, binocular cameras, trinocular cameras and surround view cameras.

(1) Monocular camera. The advantages of monocular cameras are low cost, able to identify the types of specific obstacles, and accurate identification; The disadvantage is that it cannot recognize obstacles without obvious contours due to its recognition principle, the working accuracy is related to external light conditions, and it is limited by the database and has no self-learning function.

(2) Binocular camera. Compared with the monocular camera, the binocular camera has no limitation on the recognition rate, and can be measured directly without first identification; Directly use parallax to calculate distance, with higher accuracy; There is no need to maintain a sample database.

(3) Trinocular camera. The trinocular camera has a larger sensing range, but it is a lot of work to calibrate three cameras at the same time.

(4) Surround view camera. Surround view cameras generally include at least 4 cameras to achieve 360° environmental perception.